Home

Study Designer V30 - Clinical Looking Glass from Montefiore

1. eran bellin A V Toa Be m j BE fw fm REB Gk RES Etos A i rs Arial 10 AA 7 Wrap Text General E ra tn EX A Tru zY A Paste p l BIU c Ed M amp Center 9 0 00 Conditional Formatas Cell Insert Delete Format Sort amp Find amp Format Painter 4 E Marga amp Catier TN Formatting Table Styles Clear Filter Select Clipboard n Font n Alignment n Number n Styles Cells Editing A1 i ft GroupNo A B C D E 4 1 GroupNo PatientiD MRN Disposition DRG 2 1 13177 clg 13177 REGULAR HOME SELF CARE 01 SIMPLE PNEUMONIA amp PLEURISY W CC 194 000 3 15511 clg 15511 REGULAR HOME SELF CARE 01 SIMPLE PNEUMONIA amp PLEURISY W CC 194 000 47 23278 clg 23278 SNF INCLUDES LOEB SPECIFY 03 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 5 23839 clg 23839 HOME CARE SPECIFY 06 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 Fb WP o o SPEPIHURER UBER SPEPIQAQA M o Ky Que porcus Mg one tttthsh 3 PAY 7180 Oyn ISA AAE CARE toP Ch Y 96 NANANA NA NAN SASS SIP RE PAB IMMA amp PEEDIS vM co bo 0u0 24 1 114151 clg 114151 REGULAR HOME SELF CARE 01 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 25 122955 clg 122955 AMA 07 SIMPLE PNEUMONIA amp PLEURISY W O CC MCC 195 000 26 127197 clg 127197 ICF ADULT HOME SPECIFY 04 SIMPLE PNEUMONIA amp PLEURISY W CC 194 000 27 134317 clg 134317 AMA 07 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 M Summar
2. Copy from b Select System Duration WHEN IN WITH no erga v 132 Resulting in A Edit Selected Condition __ First WBC INOT Earliest of WBC LabTestDate V WHEN IN Cohort Collection Duration V WITH No Demographics V Update and close Name MhbcFirstWhenlnCohortDuration AD JN New Analysis 13da5c4 All of Any And Save and Exit E AD N New Analysis 7da025b All of Any And Edi First WBC Earliest of WBC LebTestDete WHEN IN Cohort Collection Duration Resulting in 133 CiG Study Designer Analysis Canvas Management C0 X ra Manage Analysis al System Analysis AD B time to gastro E User Analysis USR Admit 0 365dAfterDesdWhenIn AD USR Admit_o_265dAfterDesth AD BJ usr CardiscEchaEF180DaysBefore USR Inpatient Admit Circled in the management pane is the Analysis Definition you just built Now exit Resulting in 134 d Run All Saveas h V AR X d Study Designer 5 m Wf TTO Admit W Adv TTO Admit X TTO Adv When In v BasicWbcEcholist X Reset Method Run Method Ji Show Results a uC Analysis definition a Start de No analysis definition selected Blackout 0 Days o I E i Uu Time Units Days o To Simple Me
3. F did di 162 Exit and return to Study Designer Analysis Canvas Management ox 3 gy Manage Arar I Sys B system Analysis AD earliest gastro outpt visit Choose the Analysis Definition WbcAllWhenInCohortDuration from the drop down menu Outcome Analysis definition No analysis definition selected m m Ma analysis definition selected E USR Admit O 365dAfter E USR Admit O0 365dAfterDeadWhenIn I F USR Admit 0 365dAfterDeath Bh To S USR CardiacEchaEF180DaysBefore E USR Inpatient Admit USR WbcWhenInAllCohgortDuration USR WbcWhenInFirs ortDuration Method Name All Wbc s during Method Description List Now the outcomes tab in the Available Attributes of the list populator has All Wbc 163 Study Designer P Run AII O Save As LJ V 9X x J 1 TTO Admit M Adv TTO Admit Lv TTO Adv When In KA BasicWbcEchoList lv All W Meth a Run Method Show Results etn Analysis definition Start USR WbcWhenInAllCohortDuration v Blackout 0 Days B Method Name All Wbc s During Method Description List View Wide Long Available Attributes W MOSS sL CIMYOLcinec a Selected Attributes A G 0 1 GroupNo 0 0 3 EventID M p 1 PatientID 0 0 4 EventType E 0 2 MRN 4 U Q 1 WbcWhenInAllCohortDuration X 4 IR 0 1 1 All WBC G 0 3 EventDateTime Q 0 1 1 4 EventDateTime 0 1 1 1
4. NNUNNUNNNNABB Bao c a LF NOABRWONA O XO 0 O Oi 4 0 ho 5 Xo 00 OIN Event Type InpatientDischargeDate InpatientICD9 Inpatient ICD9 Group InSet USR PneumoniaHope And Summary Summary 1 Summary 2 Data 4 The first two summary tabs provide the criteria for the two cohorts The third summary tab named Summary 2 provides the criteria for the outcomes and the last data tab provides the information of the TTO object Including Occurred whether the event happened 1 or was censored 0 Allrisk is the amount of time the patient was under observation without censorship For greater detail go to the CLG user manual Before we leave this example let s return to the criteria 31 Study Designer Fr RunAl _ Save As Lx x _ ae 4 3 B TTO admit X Run Method Demographics Target Event All Events Patient List 6 List Pivot _ Show Identifiers E fte zu n of 103 b i Total records 1646 PatientID MRN GROUPNO TimeTo Occurred Total AllRis Eun ena un EE EE You have the ability to change the units of time Study Designer 5 Run All LJ Save As NDS x x ro 4 3 TTO admit X Start Blackout 0 Days End Time from Start 365 B gt To Advanced h H Methods Groups Main By hovering over the Days the area is activated a downward arrow appears and a left click on this arrow wil
5. Run Method un zi o Outcome 5 Start End w Blackout 0 Days Time from Start 365 Admission o i r Lx S Time Units Days B To Adva nced Method Name Admit Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name Hospitalization Rate Now run 46 Hostpitalization Rate 100 90 ety i 80 60 E 5 5056 40 E T 30 20 f O 10 0 0 50 100 150 200 250 300 350 Days pvalue 0 978 BadDiabetesWRepeatGoodF1 BadDiabetesWRepeatBa Cummulative with Event 30 Days 60 Days 90 Days 180 Days 365 Days BadDiabetesWRepeatGoodFl 3 8 6 7 6 3 16 1 23 3 baseline 1 8 5 7 4 1 9 2 5 5 11 1 12 3 19 7 18 9 27 5 BadDiabetesWRepeatBadri 7 1 5 7 8 096 13 6 23 7 115 7 21 0 26 2 p In this analysis we are following patients for 365 days from their initial diagnosis and asking for the hospitalization rate during the time we are trying to bring their blood sugar under control The graph is comparing the hospitalization experience for the two groups based upon their ultimate control status e One group will ultimately be well controlled e And one group will ultimately be badly controlled We are comparing the hospitalization rate for both groups during initial time when their doctors are initially struggling to control their diabetes What do we notice In the first 365 days we see no difference in h
6. 1 xS Blackout Days End Time from Start 0 Time Units Days BP To Advanced You are given the option of entering a number for blackout days All analyses begin at the index date and in this image go forward in time to look for an outcome In some situations you do not 16 want to count any outcome in the first 2 days If that is the case you tell study designer that you want a blackout day number of 2 Study designer will then only count time and outcome that occurs after the 2 days have elapsed from each index date It is similar to resetting each cohort member s index date to date 2 days In our case we want to count admissions if they happen immediately so we will not enter any blackout day We will left click on the downward arrow in the start box to restore the image to its baseline Outcome 5 End Blackout 0 Days No outcome Ens from Start 0 selected gt Time Units Days To Advanced Start Now we will left click on the End flag Time from Start Days 1 v Start er Event Based End Date Z No analysis definition selected zx 3 H Uploaded End Dates 4 p gt Time Units Days Y B To Advanced We want to look at time to admission up to 365 days so we will enter into 1 365 days In fact we want information on 30 60 90 days as well and we will ask for that later Right now we are defining the outer limit of time by which we want to detect
7. 1 List method provides for each cohort member data on the index event demographics at the time of the index event such as age race gender and information on any other event you specify in the race track paradigm to be explained For example your cohort could have hospital admission as its index event but your list could include the last creatinine before admission or the first hgbaic in the preceding 12 months Time to outcome method calculates the amount of time elapsed from the index date of the cohort member until some target event you define For example You could have as index event hospital discharge and then target outcome event all cause hospital admission This is a readmission analytic CLG would calculate the elapsed time from the discharge to first admission and summarize this elapsed time in a graph and table Clinical Looking Glass would also count all the admissions after the discharge not just the first divide this number by the number of patient days of follow up in the cohort to calculate an incidence density of all admissions in the specified follow up time Follow up time censored by death or by running out the calendar to be described is managed automatically using the technique of the Kaplan Meier statistic Time in Range Method for continuous variables like INR or glucose CLG could follow multiple lab values for a specific lab test for each cohort member from the index date time fo
8. 61 Name AdmissionFirst 365days p NN SS S I LZ i New Analysis acb714a All 73 E Sub Group Now Left click on Within The new condition line appears on the canvas and a GUI editor appears below Als lsaveas zx x Name AdmissionFirstO 365days g ao B New Analysis acb714a All of Any And Zr d Eventi New Event Definition within 0 to 0 select time units 1 select time direction New Abstract Event Definition CINOT Aot v New Event Def v WITHIN O j TO o Days After v CohortCollection Date 62 Fill in the gui editor as follows e Name the Eventi AdmissionsAll e Leave the All of choice as is we will later choose the first admission by using Earliest at the Index Event Line e New Event Inpatient Admission you will have to build this event as you have in previous exercises e Choose time Within O to 365 Days e The entry box after After points to Cohort Collection Date This instructs the condition line to point outside of the Analysis Canvas to the cohort member s index date The resulting image in the GUI editor E Edit Selected Condition AdmissionsAll NOT All of Inpt Admit InpatientAdmissionDate W WITHIN 0 TO 365 Days After Cohort Collection Date Update Update and Close Update and Close Name amp dmissianFirst 365Days Analysis Canvas ELAD BJ AdmissionFirsto 365Days
9. 0M 0 2 1 21 Interpreting Provider Outcomes Demographics Group 0 2 1 22 Mitral Annular Calcification 0 2 1 23 Mitral Area 0 2 1 24 Mitral Mean Gradient Now drag both Count and EF Best Value to the Selected Attribute Section Study Designer ic Run All C save as y W X x P 1 TTO Admit Lv Adv TTO Admit Lv TTO Adv When In Y BasicWbcEcholist X Reset Method Run Method Outcome Analysis definition USR CardiacEchoEF180DaysBefore v ur ackout 0 Days To Simple Method Name BasicWbcEchoList Method Description List a View Wide Long B a a F 3 0 2 1 19 Echo Conclusions G 0 3 EventDateTime o9 0 2 1 20 Echo Procedure G 0 40 Disposition 2 Q 0 2 1 21 Echo Room J G 0 41 Inpatient ICD9 Group E E EEEEEELL LL SEEEEENES Sn E 5 0 1 1 4 EventDateTime 7 0 1 1 13 iLab Test Type 5 0 2 1 25 Interpreting Provider J 0 2 1 4 EventDateTime a Resulting in 150 Method Description List View Wide Lang N Available Attributes E Rad ON LFV PPPS ii SRL LLL OL at b 0 1 1 4 EventDateTime F 0 2 CardiacEchoEF180DaysBefore X i 33 0 2 1 Ejection Fraction 0 1 1 13 Lab Test Type a 0 1 1 1 Count 89 0 2 1 1 Count o F U 0 1 1 9 Lab Test Value 0 0 2 1 2 EventID ae EDO DO RE 0 2 1 4 EventDateTime 0 2 1 3
10. 229 5 Cohort 2 gt Hospitalization Index A1C gt 9 Rate The cohorts should look very familiar You built them before BadDiabetesWRepeatGoodF2 and BadDiabetesWRepeatBadF2 34 We will therefore use these cohorts in a new Time to Outcome study We will choose these two groups and evaluate elapsed time until first admission WARNING You need to think carefully about your study design You are undertaking a time to outcome study You are collecting elapsed time from some event to some outcome Which event condition line in the cohort are you referencing Edit Selected Name BadDiabetesWRepeatGoodFe t 4 Event Canvas EM INDEX EVENT Earliest of RepeatHgbAicLe7 0 And Fed DiabetesHigbA1cGes 5 Earliest of HabA1cGES 5 LabTestDate WHEN IN JunedztoJune03 WITH Age 21 to 65 1 87 RepestHgbA 1cLe7 0 Earliest of Repest LabTestDate within 180 to 365 Days After Event DisbetesHighAicGe9 5 You have two condition lines 1 Diabetic patients HgbA1c GE 9 5 2 Repeat HgbA1c LE 7 0 Your Index Event Line determines which of these two condition lines generates the Index Date Since our question is asking how much time elapses from the achievement of good or bad HgbA1c our question is asking about elapsed time from the second condition line repeat HgbA1cLE7 0 You must make sure your index event line is pointing to the second condition line i e the second event Make sure that both y
11. All of Any And zu Tn AdmissionsAll All of Inpt Admit tien The Index Event Line still needs to be changed Hover over the Index Event Line and right click Choose Edit 63 The gui to edit the Index Line appears below the Analysis Canvas Name AdmissionFirst 365Days Analysis Canvas ELAD BJ AdmissionFirsto 365Days All of Any And Lere AdmissionsAll All of Inpt Admit InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date Edit Selected Condition AnalysisDefinition All of Any Root Operator AND Update Update and Close We need to change the Index Event Line from All of to Earliest of Earliest at the Index Event will choose the first admission A Time to Outcome Method must go to a singularity a single event per person An Analysis Definition used in a Time to Outcome Method must go to either an Earliest or Latest Event It must go to a singularity not to All Only Analysis Definitions with Earliest or Latest not All of in the Index Event Line will show up in the Analysis Definition drop down menu in the advanced mode Time to Outcome 64 Make the modifications in the GUI editor Edit Selected Condition l Analysi sDetinition Alof Any CE en Operator AND a AdmissionsaAll e Choose Earliest Of e Change Any to focus on the conditi
12. Count 0 1 1 13 Lab Test Type 0 1 1 2 EventID 0 1 1 3 EventType Outcomes Demographics Group 0 1 1 4 EventDateTime 0 1 1 5 EventName 0 1 1 6 Categoric Lab Test Value 4 n3 37 Greiuined Numeric ahe P The default is wide view 164 d Run All Saveas X X j Study Designer i E X TTO Admit WV Adv TTO Admit XV TTO Adv When In V BazicWbcEchoList All We a C Reset Method Run Method un e tit Analysis definition e USR WbcWhenInAllCahartDuration Uu Time Units Days Methods B To Simple The default is Method Name All Wbe s During wide view Method Description List View Wide Long FRY ST 2 KM KA 4 0 Common Event Attributes U 1 iGroupl 0 0 1 EventName 0 0 2 EventDateTime e 0 0 3 EventiD A G 0 3 EventDateTime 0 0 4 EventType 0 1 1 4 EventDateTime 4 JA Q 1 WbcWhenInAllCohortDuration X 0 O 1 1 13 Lab Test Type 4 EX 0 1 1 AII WBE omes Demographics Group Restore Default aa a Obed EE xd esue View Wide Long g The default Wide view will create a separate column for each wbc for each patient If even one of the patient s in the cohort had 100 wbc tests then you would have 100 columns for wbc values and 100 columns for event date for every patient Even if the next patient in the cohort had only one WBC he would also have 100 columns although t
13. Run Method Outcome Analysis definition B To Simple Method Name Ady TTO admit Method Description Time To Outcome rr Tt F F F Estimation Points 30 60 90 180 365 Days Graph Name KM Graph Name 60 This opens the Analysis Canvas where we will build our Analysis Definition AD CiG Study Designer Analysis Canvas mai Management C9 X E __ _ l Edit Selected C Save As a Manage Analysis A Name New Analysis acb714a zi sys System Analysis K CAD D earliest gastro outpt visit f Analysis Canvas ewgso LIN New Analysis acb714a All of Any And H a User Analysis 1 New Analysis acb714a 3 asarmien ADG_A1C POC CPT_Sep2014 asarmien ADG_A1C_POC value_Sept2014 USR Albumin30_210befLast NS Let s build an AD to imitate the simple method with e Elapsed time O to 365 days e The Earliest Admission after the cohort index event date Like all canvases you hover over the Index Event Line and right click IZIS savas z z9X X Name AdmissionFirst0_365days x A Analysis Canvas m ras New Analysis acb7 14a All of Any wt You are given the option to Add a condition to the canvas or to Edit the index event line Choose add a condition Since we want to compare admissions that occurred within 0 365 after hospital discharge the new condition line on the canvas will be Within
14. AL te xo Hover over the methods tab and left click 116 Save s y TX x Group Mame Cohort Collection 1 Mele nemora c uss 9neumc is Males 2010 Discharge Primary CE Female 2neumcois cusa Pneumonia Females 2010 Discharge Primary E Click the sign to add another method d Study Designer d Run All L Save As m X 3 TTO Admit X Adv TTO Admit iv TT Show Criteria Target Event All Events Patient List Choose List from the drop down Name the Method BasicWbcEchoList Left click Add 2 Adw TTO Admit Time to Oosftcomae Time To Outcome TTO Adw When In Time to Ostcema Time To Gutcome repre E New Method Dialog x Method Type Select a Method Type New SU Method Name 5 Time to outcome List Time in Ra wat The image below appears Add i Cancel Run Run 117 d Study Designer d Run All Saveas V X a em Y TTO Admit Wf Adv TTO Admit 4 TTO Adv When In BasicWbcEchoList X mj A ul o PU ome 5 End 3 No outcome Time from Start 0 E n selected T Ib Time Units Days To Advanced Method Name BasicW bcEchoList Method Description List Ld TEE x View 4 Wide Long ms b iew Wide Long E 4 Available Attributes Selected Attributes P E E a 4 G 0 Common Attributes G 0 1 GroupNa 5 G 0 1 GroupNo J
15. Admit Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name YI SP Save your study 88 Next we will build an AD with death as an event that occurs during the durational event of the admission Approach 2 You will be building an AD with two lines The first line will look for an admission within O to 365 days of the cohort collection This admission will not have a dispositionzexpired We will not require any disposition at all The second line will have an event death when in the event of the first condition line on the Analysis Canvas The when in condition line requires a death during the first admission following the cohort s index date Condition Line one Earliest Admission within O to 365 days of the cohort collection And Condition Line 2 Earliest Death In House or Social Security When In during the duration of the admission Add a new method in Study Designer Study Designer 1 Studies 3m Study Designer Run All O save As a X g s USR Admission Repeat Diabetic amp ls USR Diabetes control outcome 5 USR Pneumonia g s USR Pneumonia Readmit P ES Mens x W TTO Admit 4 Adv TTO Admit X Main Jl Show Criteria Demographi Target Event All Events Patient List 100 i 90 T Groups Methods Choose a Time to Outcome method 89 New Method Dialog x Method Type kei Me
16. All of Any And ie Eventi New Event Definition within 0 to 0 select time units select time direction New Abstract Event Definition Edit Selected Condition Ejection Fraction PINOT Earliest of Cardiac Echo EchoDateTime WITHIN 0 TO 180 Days After Cohort Collection Date ss Update Update and Close Update and close EH AD BJ CardiscEchoEF180DaysBefore All of Any And Leert Ejection Fraction Earliest of Cardiac Echo EchoDateTime within 0 to 180 Days After Cohort Collection Date Save and Exit 145 analysis Canvas Ap B CardiacEchoEF180DaysBefore All of Any And L S my Ejection Fraction Earliest of Cardiac Echo EchoDateTime within 0 to 180 Days After Cohort Collection Date View in the management pane the new AD called USR CardiacEcho180daysBef Ab Manage Analysis Sys B system Analysis fuser j user Analysis AD J USR cardiacEchoEF180daysBef AD B USR wbcrirstWhenInCohortDuration Now exit to return to Study Designer criteria page Analysis Canvas 146 From Analysis Definition dropdown menu choose CardiacEchoEF180DaysBef 147 You have the racetrack paradigm exposed Start Blackout 0 Days Add the Analysis Definition by clicking on the black downward arrow and left clicking aan d Analysis de
17. C we see a count on each row equal to 1 meaning there is only one observation per row Column D are the actual white count values 167 Next Steps In Class exercise Gatifloxacin see handout Does it cause dysglycemia Review the help manuals Review the videos 168
18. Canvas Webpage Dialog ELOPED i SKILLED NURSIMG FAC SHORT TERM SUB ACUTE 4 a https clg emerginghealthit com CL GNET Madules EventDe REGULAR HOME SELF CARE PRIVATE ASSISTED LIVING P Edit Options HOME CARE NON HOSPICE PERSONAL CARE amp HOM STILL PATIENT OR EXPECTED TO RETURN ACUTE CARE PPS HOSPITAL SPECIFY EXPIRED IN MMC HOSPICE UNIT EXPIRED PLACE UNKNOWN FEDERAL HOSPITAL HOSPICE HOME CARE HOSPICE FACILITY INPATIENT ACUTE REHAB includes LUBIN LONG TERM CARE HOSPITAL PSYCH HOSP AND DISTINCT PSYCH UNITS SNF INCLUDES LOEB SPECIFY ICF ADULT HOME SPECIFY OTHER HEALTH CARE INSTITUTION CANCER CENTER OR CHILDREN S HOSPITAL HOME CARE SPECIFY Please Select PENDING INFORMATION ADMITTED TO MMC HOSPITAL SPECIFY OTHER HOSP FOR NEONATE WGTGAIN AER EXPIRED Event Type inp Name n p atie ntA dmitWD eath Unite se Class Class F Event Definition Options nfm Undefined Condition NOT Disposition Please Select Update Update and Close This results in Fr lt f m A Definition vas Webpage Dialog 2 Definition Canvas Dag e https clg emerginghealthit com CLGMET Modules EventDefinitionBuilder EventDefinitionBuilder aspx Def l ypez EVEMT amp Sellv End Point E Event Definition Options E Definition i Rep Undefined Condition EXPIRED Update Update an
19. EventType G 0 2 1 23 EF Best Value CE LI 2 1 4 Eventbatelime 0 2 1 1 Count al t 1 3 3 5 FventName omes Demographics Group Now run BasicWbcEchoList Study Designer P Run All L Save as p el gt X X 4 F ES v TTO Admit y Adv TTO Admit v TTO Adv When In BasicWbcEcholist X Show Results Main Outcome Analysis definition USR CardiacEchoEFi180DaysBefore Start Groups p Blackout 0 Days o L J 1 E tment dys 8 Method Name BasicWbcEchoList Method Description List Save View Results 151 35 00 20 00 20 00 65 00 65 00 60 00 70 00 65 00 2 24 2010 8 11 00 AM 12 6 2010 10 51 21 AM 8 19 2010 11 41 00 AM zl nmm TTO Admit WV Adv TTO Admit TTO Adv When In BasicWbcEchoList X 0 ER Reset Method Run Method Show Criteria w a iz List Pivot Show Identifiers Export Kd 4 Page of 106 b p Total records 1683 E me OutcomeCardiacEchoEF180DaysBefore Ejection Fraction 1bEF Best Value OutcomeCardiacEchoEF180DaysBefore Ejection Fraction 2EventDateTime OutcomeCardiacEchoEF1800 60 00 35 00 50 00 All cohort members are represented with a row You will note that some rows have no ejection fraction Those without an echo are left blank on the outcome for Ejection fraction 152 List Method asking for all results not just the earliest or l
20. First Admit rTYrYVYY YT A AAAAA Run the Method You can now compare in the Target Event Tab between methods TTO admit and TTO Advanced and find that the results are the same This means that the Advanced Mode when replicating the criteria of the Simple Mode gets the same result 70 Target Events Tab in the Simple and Advanced Modes are identical notice the highlighted tabs Readmisasce for M F Pneumonia Pte Ade TTO Admit 0 s w 60 a 2 00 20 o 5 Q w d 325 023290 00 39 Days tuus 0575 Days paue 0575 Wwe Pseuona Fame Pneumonia uno Demographics between these two methods are the same Notice Advanced TTO Atm is hihlighe Noice TTO Admitin the Simple 4 n Y TO Admit Y Adv TTO Admit X Mode is highlighted p Run Method Show Criteria Demographics p Bent Al Events Patient Lit Male Pneumonia 754 100 0 0 0 0 Female Pneumonia baseline 0 0 929 100 0 0 Male Pneumonia 754 100 0 0 U rir 0 0 929 100 0 0 Male Pneumonia 0 098 1219 2250 22 56 75 85 89 9 n 0 109 206 3 42 62 79 S8 9 1M MalePneumonia 0 099 1210 220 22 56 75 8 go 9 Female Pneumonia baseline 0109 206 3 42 amp 7 8 9 104 71 All Events tabs between these two methods are the same Jemographics Target Event All Events Patient List Number Target Events Achieved Person Days at Risk Incidence Density per year of ri
21. House or 55 All of DesthInHaussOrSS SSAorinHouse WHEN IN Admit 0 355dARer Ing You now see the new AD in the Analysis Canvas Management Pane Analysis Canvas Now we want to use our AD in our Time to Outcome Analysis Return to Study Designer by left clicking management pane s close button 101 Analysis Canvas i Management ox aD Manage Analysis Svs B system Analysis AD Bj eriiest gastro outpt visit AD B time to gastro Userg User Analysis AD Bj usn Adm 0 365dAfter Now we find ourselves back in Study Designer Advanced Mode Study Designer A Run All _ Save As gt x x TUT V TTO Admit Adv TTO Admit TTO Adv When In X Main Run Method Outcome Analysis definition No analysis definition selected Groups unm I a I E F Time Units Days e To Simple Method Name TIO Adv When In Method Description Time To Outcome Estimation Points 70 60 90 180 365 Days Graph Name KM Graph Name TF FFT F F Now we want to select the Analysis Definition we just built from the dropdown menu by left clicking on the downward arrow 102 Study Designer P Run All C save as V mo X Methods Em V TTO Admit af Adv TTO Admit TTO Adv When In X Run Method Outcome Analysis definition No analysis definition selected No analysis definition selected E F USR Admis
22. K mm rd Meee l ae Mane Left click on the Analysis Definition and choose No analysis definition selected j Study Designer d Run All C save as X i V TTO Admit WV Adv TTO Admit X TTO Adv When In V BasicWbcEcholist X Reset Method Run Method Show Results ch 3 tut Analysis definition 5 Start y Bul anaiysis definition selected v Days No analysis definition selected E I USR Admit 0 365dAfter t USR Admit O0 365dAfterDeadWhenln USR Admit 0 365dAfterDeath B To USR CardiacEchoEF180DaysBefore USR Inpatient Admit Method Name BasicWbcEchoList USR WbcWhenInAllCohortDuration le le im Te im Im Im 4 USR WbcWhenInFirstCohortDuration Method Description List ie TT T T Open the canvas using the triple dotted button 141 pnm Analysis definition Start Blackout 0 Days o Time Units Days p To Sim ple Method Name BasicWbcEchoList Method Description List A new Analysis Canvas opens Analysis Canvas l Management Bias F mapas Save As a gt x x RD Manage Analysis Name New Analysis 7787263 sl Sws B System Analysis AD B aes cedi E edt analysis Canvas AD B tme to gastro AD H New Analysis 7787263 All of Any And ser user Ansiyss Name your Analysis Definition CardiacEchoEF180DaysBefore Save As ll al X F analys
23. Meitsum Sth 10th 25th 16 Choose the Target Event tab Study Designer p3 Run All C save as X X Main Methods Groups 4 s amp ES v Admission X Run Method Jl Show Criteria Demographics Target Event All Events Patient List 100 t 90 Ga it 8056 7056 60 e E 50 40 30 20 O 10 0 30 100 150 200 Days BadDiabetesWRepeatGoodF2 baseline 21 36 39 45 52 58 62 64 BadDiabetesWRepeatBadF2 21 31 36 43 51 58 62 64 65 65 Hospital Admission Explore the resulting tabs Demographics TargetEvent All Events Patient List 290 300 350 pvalue 0 00213 BadDiabetesWRepeatba 41 Cummulative with Event 30 Days 60 Days 90 Days 180 Days 365 Days BadDiabetesWRepeatGoodF2 2 7 4 6 9 n 5 4 9 1 15 3 baseline 4 3 17 6 6 2 12 0 11 6 18 9 BadDiabetesWRepeatBadF rm o4 3 P 7 47 i r D E E F 2 9 5 799 8 5 14 39 23 09 Figure 1 Elapsed Time to readmission from moment of Good LE 7 or Bad GE 9 Control We note the Diabetic patients with a bad repeat HgbA1c test are climbing into the hospital faster than the Diabetic patients who had a good repeat HgbA1c The cumulative incidence of first admissions for Diabetics with a bad repeat test at 365 days is 23 while for Diabetics with a good repeat test it is 15 3 The p value is 002 and thus highly significant The 95 confidence interval of the two estimates at 365 da
24. Options qzheng death Definition ebellin7 ed triage Note this canvas event definition builder does not have any time element and does not have any quantity all Earliest Latest The reason for this is that the Event definition is used in the simple version of Study designer where the racetrack paradigm imposes the notion of earliest on the event and the time is provided by the user in the end flag Left click on Save and Exit 18 The management pane now records the admission event Event Definition Builder Management i Manage Definitions th Event Definitions ebteacher Admission Now Left click on the close button at the top of the Event Definition Builder management pane and return to the study designer Left click on the downward arrow next to No outcome selected Its s Event Definition Start No outcome selected lt Jl v Blackout 0 Da Admission X lt admit with Exclusions ime Units Days Be death ed triage id Name TTO admit mortality id Description Time To Ouredmme Left click on Admission OSs Event Definition Start Blackout 0 Da Time Units Days To Advanced Hover over the outcome s box downward arrow turning it yellow and then left click on this downward arrow Yielding 19 ESSET UE J RunAU O save as f ld me Eip TTO admit X Run Method Uu a r o Ou
25. Start 0 ra selected d T x Time Units Days B Ta Advanced Method Name Basic WbcEchaList Method Description List View Wide Long 4 T CT CN oh Available Attributes de E a 4 G 0 Common Attributes D D 15 Gender k G 0 1 GroupNo D D 20 Race u G 0 2 EventName D D 22 Age at Index Date m G 0 3 EventDateTime Fo G 0 3 E i 3 0 3 EventDateTime 0 3 EventDateTime e E A G 0 4 EventID ih G 0 40 Disposition e mE e G 0 5 EventType ia G 0 41 Inpatient ICDS Group a E y G 0 6 AdmitToInpatFromED Order Dateti G 0 13 APR DRG z l Restore Default Study Status The racetrack is in the default Simple Mode We want to change the racetrack metaphor to the advanced mode Left click on toggle switch To Advanced ponon cd Anales definition i tat hM Fi analysis definition selected End in from 5pm 8 Blackour Cays Start Binckour U Cars Notice the toggle switch in the Advanced mode says To Simple meaning that if you want to switch from the Advanced mode back to the Simple mode click here We now want to get the First White Count for each patient in the index hospitalization We need to build an Analysis Definition that will find the first white count for each patient in the index hospitalization To build an Analysis Definition left click on the three dotted button 124 UT anam definitions tt tst S Start A
26. a readmission In this intro course we will not cover 2 3 4 However these controls give you exquisite ability to censor follow up time by an intercurrent event event based end date represented by an analysis definition or by an uploaded end date Close the end arrow by left clicking on the downward arrow 17 Start End Blackout 0 Days O outcome Time from Start 365 Time Units Days B To Advanced Now we have to define the outcome of interest Hover over the Outcome box and note how it turns active by changing color to yellow Left click on the Outcome Box teu Event Definition v 3 No outcome selected B uu Start Blackout 0 Da ZG mm S Time Units Days BP To Advanced The box opens up and gives you two choices 1 If you already have a defined event definition from a previous study the system will let you left click on the downward arrow and select it 2 f you do not already have a defined event definition the three dotted button opens up a canvas and allows you to build the outcome event Let s Left click on the three dotted button and see the canvas and build an admission event CG Study Designer Event Definition Builder i Management iE save As Manage Definitions Event End Point Name npatient Admit Inpatient Admit v Please Select v lt Ebb Event Definitions p _ Type pa Event qzheng Admission i ebellin7 admit with Exdusions Event Definition
27. build an Analysis Definition that tells Study designer to follow elapsed time from O to 365 days after the index event of the cohort member Until the first hospital admission with a disposition of death 80 AESESEEESE T saveas 2 X X Name New Analysis 3b62bb analysis Canvas AD H New Analysis 3b82bb7 All of Any And Type the name Admit O 365dAfterDeath to your analysis definition Analysis Canvas AD M New Analysis 3ef79cb All of Any And Right click on the Index Event line and add a Within condition Edit Selected CT Save As fa Ga Admit_0 365dAfterDeath Name p Analysis Canvas Name you event definition Admit O 365dAfterDeath Choose All of Choose time 0 to 365 Days Build death into an inpatient admission by using an attribute of the admission disposition expired by doing the following 900 81 Highlight New Event Definition and left click Edit Selected Save As e x X ELAD M New Analysis 3ef79cb All of Any And tae Eventi New Event Definition within 0 to 0 select time units select time direction New Abstract Edit Selected Condition Admit_O_365dAfterDeath WITHIN 0 TO 365 Days After CohortiColl Update Update and Close Select New Event Def from the drop down menu New Event Def Select from Edit Selected Conditio Select from Common Defintion gt PO Admit 0 365DaysAft
28. method Time to Outcome Name the Method Adv TTO admit then add Method Type Time to outcome eel Method Name Adv TTO Admit 58 Notice that the new method is shown as a tab on the top of the Study Designer canvas We now see two methods in the study You see TTO admit and Adv TTO admit Study Designer L SaveAs X X E e a TTO admit Adv TTO admit EL nr Two methods in this __Run Method study a o ul n uh No outcome Time from Start O 5 l selected z eee aS qi z P Method Name Adv TTO admit Method Description Time To Outcame i Estimation Points 30 60 90 180 365 Days Graph Name KM Graph Name We now want to move from the Simple Method To Advanced Method Hover over the To Advanced Button The button becomes yellow Left Click on To Advanced The racetrack image now changes to the Advanced mode 59 Outcome Analysis definition 1 Note how the button that used to be To Advanced now becomes To Simple This button is a toggle switch Click on it a few times until you deeply appreciate what is happening 2 Consider the rightmost flag the Analysis Definition Recreate the same analysis in the advanced mode that we developed in the simple mode Right click on the three dotted button see image below Study Designer amp Run All 1 CJ save As Ll Ljex xo 4 Ea y TTO admit Adv TTO admit X
29. saves the results and the rules and must be done after every run 21 Now let s look at the results SUVs dy Run All save as V X X 4 TTO admit X 1 Run Method Show Criteria 2 Main Demographics Target Event All Events Patient List se Ge a r 741 100 0 0 0 0 Female Pneumonia 0 0 905 100 0 0 e Minimum Sth 10th 25th 50th 75th 90th 95th 100th Male Pneumonia I line 0 098 1 207 2 232 22 57 75 85 89 99 Female Pneumonia 0 109 2 116 3 42 62 79 89 94 104 Methods Groups After the run you continue to see the 1 TTO admit Method tab 2 And now next to the Run Method Tab is a Show Criteria Tab This is a toggle switch between Show Criteria and Show Results Clicking on Show Criteria reveals 22 ET Designer Run All O Save As i3 jer x pr TTO admit X 5S m L 2 Start Blackout 0 Days End Time from Start 365 Admission To Advanced Method Name TTO admit Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name Readmission for M F Pneumonia Pts Click on Show Results to get the results again E 1 2 V TIO admit X L Show Criteria i 2 l O Demographics Target Event All Events Patient List Male Pneumonia baseline Female Pneumonia 0 0 905 100 0 0 741 100 0 0 0 0 Age Percentiles Male
30. 180 365 Days Graph Name KM Graph Name cr wor FF We will use the triple dotted button to move into the Analysis Canvas and edit our Analysis Definition This opens the Analysis Canvas Analysis Canvas Fog Save As Manage Analysis Name New Analysis 7c4edc1 E Gyn Aredusss L1 earliest gastro outpt visit E Analysis Canvas LAD B time to gastro AD Pi New Analysis 7c4edc1 All of Any And iserBl User Anais Our plan is to modify a preexisting Analysis Definition Therefore we have to exit the right hand side and go to the management pane Click on Exit 92 Management Bae ML Etech I saveas uz X is Manage Analysis Name New Analysis 7c4edc1 a System Analysis r Analysis Canvas Click on AD B time to gastro AD M New Analysis 7ctedc1 All of Any And X serll user anzivss Now for some housekeeping In the management pane look for the incomplete Analysis Definition Analysis Canvas A Management PARAS i Sys B System Analysis LAD JJ esriest gastro cutpt vist As you can see the AD called New Analysis 7c4edc1 has a right white stripe not green stripe The lack of green means that it is an incomplete AD Now delete this AD 93 Without this AD the management pane looks like Analysis Canvas Management DEZ Ati Manage Analysis Svs System Analysis LAD JJ earliest ga
31. Before USR Inpatient Admit Method Name BasicWbcEchoList USR WbcWhenInAllCohortDuration Method Description List it inn lemons Available Attributes Selected Attributes a 0 0 Select Outcome to display attributes G 0 1 GroupNo f p 1 PatientID 3 p 2 MRN 3 3 ses D 14 Ethnicity D 15 Gender 0 19 Preferred Language 0 20 Race 3 0 22 Age at Index Date G 1 1 5 EventDateTime G 1 1 39 Disposition Resulting in V 8 wde Ctm qi CILIIDTER M 00 Common Event Attributes GALA Group P 4 M oa WbcWherinhrstCohortDuraton M 30a Patenti 4 BR 0 1 1 Fest wer oz wan b 0 1 1 1 Count Do ses 0 1 1 2 EventiO 3 0 14 Ethnicity 0 1 1 3 EventType 0 15 Gender DOLLA FvevtDateTime 3 0 19 Preferred Language 0 1 1 5 venthiame 3 0 20 Race 0 1 1 6 Categorc Lab Test Value 3 0 22 Age ot Index Oste 0 1 1 7 Greuped Numeric Labs A GLAS FventOateTime 0 1 1 8 Lad Text c 1 1 39 Dispositon 136 Note EventDateTime in Available Attributes for the White Count AD is grayed out meaning that we should expect to find it under Selected Attributes when we move the vertical slider View Wide Long N Selected Attributes 3 gt Q 0 Common Event Attributes 3 19 Preferred Language OO A4 U 0 1 WbcWheninFirstCohortDuration X 73 p 20 Race i 4 BR 0 1 1 First WBC 3 p 22 Age at Index Date E 0 1 1 1 Count G 1 1 5 EventDateTime E 0 1 1 2
32. CLINICAL LOOKING GLASS Study Designer Clinical Looking Glass A Self Study Guide Eran Bellin Version 30 October 27 2015 Contents Precs ASSENTOS ea E m 3 OV DE ENET r EE E E 3 AOMC MEDO n ETE 5 KE CO OUT COMMS e a E S 7 Exercise 1 Pneumonia Readmit SEHOV scicssssacssatsveresedsunicscteeressedbecdenesisannsedaeetuandevaroned tae teiesteweeadseniorderernans 9 FSIS Sr VO gi ED a E E A E I OiMiKN MEAS 34 Tutorial 16 Importance of Cohort Index Date on the meaning of a Study eeeeeeeeee 44 Time to Outcomes Advanced Method oseiess corio ERN eUE ove a XM HV IN TER PrcUt Va UNE C epdeuruceenceveiacenamiess 50 First Recapitulate Results of Simple Method with Advanced Method eene 50 Complex TTO Examples requiring an Advanced TTO ccccessccccesseccceesecccceesececeeeceeeeeeceesauecessegeceeseeeeeeeas 74 Summary Utility of Advanced Method TTO ccccccssccccesseccceeseccccesececeesecesseeceeseeseceseuaecesseeceeseeneeeeas 106 List Method Looking only at the attributes of the cohort esessssesessensesserssrrrnesreresrrereseeresrressrrereseeresseeee 107 List Method When you want to see attributes that were not used in the building of the original cohort E E 115 List Method asking for all results not just the earliest or last esseesen 153 Tutorial 24 List Method Long view cccccessccccessccccesececeesececeeeseceeeusececsueceeseese
33. D omes Demographics Group Aen 3 evant uns Restore Default Run the method View Wide Long H 13 D 1 PatientID M D 2 MRN G 0 3 EventDateTime 0 1 1 4 EventDateTime Q 1 1 13 Lab Test Type 0 1 1 9 Lab Test Value 0 1 1 1 Count a 166 We export to an excel spreadsheet and have eliminated some columns Resulting in X id H i 27 db84f ca06 4992 b132 b36221c2e2fa Read Only Microsoft Excel File Home Insert Page Layout Formulas Data Review View Add Ins Arial ho la a P General ES Conditional Formatting 7 ga Insert 7 aa 7 E E Format as Table gq Delete 7 jns iod bs EE Oo amp SSS od io 3 0 E Cell Styles 7 Format Clipboard Ma Font P Alignment E Number z Styles Cells E1 1 jx EventDateTime C D PatientID OutcomeWbcWhenlnAllCahartDuration All WBC Count QutcomeWbcWhenlnAllCohortDuratian All WEC Lab Test Value 13177 1 48 13177 1 A 13177 3 7 1317r 1 j4 13177 1 21 15511 1 10 7 15511 1 g 15511 8 3 23218 1 4 23218 1 52 23278 1 Af 23218 1 4 23218 1 4 1 23278 1 52 232718 1 54 23218 1 13 2a218 1 8 2 23639 1 58 23839 1 54 23839 1 T 1 23839 1 72 23839 1 85 23839 1 T5 23839 1 8 7 24380 1 13 2 24380 1 10 8 24380 1 95 5 24380 1 8 5 241794 1 11 8 Some comments Column B with patient ID demonstrates that some patients have multiple rows This means that they have multiple Wbc values from their hospitalization available for review Column
34. D 1 PatientID g J 4 G 0 2 EventName D 2 MRN m 5 G 0 3 EventDateTime 3 n 3 ses a E A G 0 4 EventID 3 D 4 Apartment a al G 0 5 tEventType D 5 city a E A G 0 6 AdmitToInpatFromED Order Datet J D 6 State i i Restore Default ie pe ee oe a Study Status 4 First we are going to concentrate on those attributes inherent to the cohort member s index event The attributes used in building the cohort are automatically moved from the Available Attributes to Selected Attributes Those attributes that were not included in the conditions used to select cohort membership have not been brought over A number of demographics are automatically brought over We are only interested in keeping SES Gender Race Ethnicity Age at Index and Primary Language We must remove the rest Click on List populator maximizer 118 View e Wide Long 5 Left click on the Demographics tab i E93 M TTO Admit Adv TTO Admit TTO Adv When In BasicWbcEcholist X Reset Method Run Method u o Start i un Blackout 0 Da Time from Start 0 B li ys selected r 7 D Time Units Days BP To Advanced Method Name BasicWbcEchoList Method Description List TN om iew Wide gt Long HB Available Attributes Selected Attributes N A D D 8 CensusBlock 3 D 9 Latitude D D 10 Longitude D D 11 CensusBlockGro
35. D 6 State m i IE G 0 7 APR DRG Code D 7 CensusTract G 0 8 Inpat Account O D 8 CensusBlock G 0 9 DualEligible 3 D 9 Latitude G 0 10 Admit Referral Agency C D 10 Longitude G 0 11 Discharge Referral Agency D D 11 CensusBlockGroup Restore Default 4 Study Status 4 Available attributes that can be moved from 4 to 5 are now very visible 1 Group attributes are visible when you left click on Group 1 within the BasicListforCohort Method Grayed out attributes in the Available attributes section 4 have already been selected and appear in the selected attributes section 5 To select new attributes Left click on the Available attribute to select more than one available attribute hold control key down while you left click on multiple attributes then drag to the selected attribute pane on the right Drag DRG to the right 109 Study Designer amp Run All _ Save As x i i icListF x E 4 E TTO admit Adv TTO admit BasicListForCohort Reset Method Run Method a z View Wide Long m pen OQ Available Attributes Selected Attributes x a F G 0 1 GroupNo 3 G 0 10 Admit Referral Agency 0 1 GroupNo c 5 ars 2 G 0 11 Discharge Referral Agency 1 D 1 PatientID el 3 2 MRN S G G 0 13 Drg Weight T 311 a SES E 6 0 14 Admit Attending D 4 Apartment a 7 G 0 15 Admit Housestaf
36. Def Left Click on New Event Edit Selected Condition First WBC NOT Earliestof Select from Select from Common Defintion Copy from 127 Ww amp Event Definition Palette Webpage Dialog e https clg emerginghealthit com CLGNET Modules EventDefinitionBuilder EventDefinitionBuilder aspx DefType EVENT amp SelMenuOptionz NewEventDefinition amp EntityID ew i Name EV Ced19 Event Please Select End Point Event Please Select Type l Event Definition Options Name the Event Type WBC Event oossoo Name WBC Type Please Select End Point Event Please Select update Event Definition Options Choose Lab Test from the dropdown menu Name WBC Spei Please Select End Point Event Please Select Lin Please Select ST Ambulatory Surgery Admit Ambulatory Surgery Discharge Assessment Functional ADL Assessment Functional IADL Cancer Registry First Recurrence Date Cancer Registry Initial Diagnosis Date Cancer Registry Site Spec Factors Init Dx CPT4 Procedure Date Date of Birth Death In House Death Date Death Social Security Death Date Death Social Security or In House Death Date Echo Date ED Discharge ED Seen ED Triage Findings Immunization Date Inpatient Admit Inpatient Discharge Internal MedAdminDate ab Test Med Orde Start Med Order Stop Event Defini
37. Description List View Wide Long m Available Attributes Selected Attributes a a a 3 4 G 0 Common Attributes G 0 1 GroupNo G 0 1 GroupNo J f D 1 PatientID z G 0 2 EventName 33 D 2 MRN s G 0 3 EventDateTime 3 p 3 ses 3 A G 0 4 EventID 3 D 4 Apartment T G 0 5 EventType bs city amp G 0 6 AdmitToInpatFromED Order Datetime E ID D 6 State q To simplify this exercise we are going to delete all of the Selected Attributes chosen by default by Looking Glass Click on the list populator maximizer l Study Designer P Run All L Saveas k n X X 4 37 TTO Admit Lv Adv TTO Admit Ki TTO Adv When In Lv BasicWbcEchoList All W Main Reset Method Run Method Groups 3 4 G 0 Common Attributes E G 0 1 GroupNo t o G 0 1 GroupNo D 1 PatientID Z Acozievenname DLL OMNEM Lud A G 0 4 EventID 3 D 4 Apartment T A G 0 5 EventType 3 p s city 5 G 0 6 AdmitToInpatFromED Order Datet 3 D 6 state 8 G 0 7 LengthOfStay Analytic days 3 D 7 CensusTract A G 0 8 LengthOfStay days 6 D 8 CensusBlock A G 0 9 Admit Nursing Station 3 n o Latitude G 0 10 Admit Priority p 10 Longitude G 0 11 Admit Referral Agency 3 D 11 CensusBlockGroup A G 0 12 Inpatient Admit Service Group 3 D 12 Patient Address A G 0 13 APR DRG 3 p 13 Date Of Birth G 0 14 APR DRG Average LOS 3 p 14 Ethnicity G 0 15 APR DRG C
38. Designer Now we have to tell study designer to switch to its Method canvas so we can paint the rules for the method we are about to select Left click on Methods tab dy Runan O save as V X X E un 2 o1 un E 482 4 LS 4 4 ad We are now on the Methods canvas of the Study Designer Note how the Methods tab 2 is now white active and the Groups tab is now grayed out The Methods canvas has no method yet We will add our first method by left clicking on the sign Please note we will be able to add many methods on the method canvas but of course we will start with our first 12 Study Designer C save as blimps X New Method Dialog Add Method Type Select a Method Type z New Sti Method Name Method 1 Www Ww Ww WW M M MM ARA Add Cancel ud A dialog box opens up allowing us to choose from the drop down menu from the three method types available to us and allowing us to name the method Naming is important if you are going to have many methods in your study Left click on the downward arrow next to select a method type We will choose Time to outcome from the method type and we will name the study TTO admit in recognition of the fact that we are going to evaluate the amount of elapsed time to first readmission from live discharge of male and female pneumonia patients in 2010 New Method Dialog Add Method Type Select a M
39. EventID G 1 1 39 Disposition 7 G 1 1 40 Inpatient ICD9 Group 3 G 2 1 5 EventDateTime E G 2 1 39 Disposition 0 1 1 6 Categoric Lab Test Value 0 1 1 7 Grouped Numeric Labs 0 1 1 8 Lab Text 1 E Restore Default We want to drag Lab Test Value and Count Lab test value will provide the actual White Blood cell count Count will provide an indicator variable for the count of results Since in this case we are looking for earliest count will be either O or 1 Since in this case count is referring to a singularity a count will be either O or 1 Maximize the list populator de Run AM savea VQ X X a Ea 4 TTD Admit M Adv TTO Admit M TTO Ady when In Di BasicWheEcholist X Reset Method Run Method Show Results F na us dehnibionz E USR WbcWhenInFirstCohortDuratis wn Blazicz t 0 Dari n x a mme p Te Semple list populator Method Name BasicibcEchoList Method Description List Vine Wide Leng Available Attributes F 0 113 EventType 1 3 0 22 Age at Index Date Ss i 1 1 4 A G 14 5 rEventDateTime i 8 0 1 1 5 Eventame if G 1 1 39 Despositian s 0 1 1 6 Categoric Lab Test Value i5 G 1 1 40 Inpatient ICD9 Group i E 0 1 1 7 Grouped Numeric Labs j G 2 1 5 EventDateTime 8 0 1 1 8 Lab Text iA 6 2 1 39 i Disposition 137 Drag over Count and Lab Test Value from Available Attributes into Selected Attrib
40. O0 365dAfterDeadWhen i B To Simple Method Name TTO Adv When In Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name DeathAfterReadmissionTT 104 Remember to Save once the method has run d Study Designer D Run All O Save As i mE S x 4 s Em y TTO Admit r4 Adv TTO Admit M v TTO Adv WhenIn X Run Method Show Results Main uaria Analysis definition Groups Start un Blackout 0 Days I E F Time Units Days Method Name TTO Adv When In Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name DeathAfterReadmissionTTO rT rr cw cw Review the results from the demographics target events and all events tabs Demogn pics Target Event All Events Patent List CeakhAftergeadmiscsonTTCO 105 Dass puse 0 138 5 Mala Peeumonia Famale Pneumonia p osnegagies Target Event AI Events Patient Lot a Number Target Hale Pneumonia 74 243062 Bgm m RR 126 306170 0 15 Summary Utility of Advanced Method TTO 1 You can have latest event or earliest event while the simple method permits only the earliest event 2 Youare able to have complex outcomes with multiple condition lines on the analysis definition while in the simple method you can have only one event as an outcome 3 Tutorial 19 Meaning of Earliest vs all ppt 6 https www youtu
41. Pneumonia jo baseline 0 098 1 207 2 232 22 57 75 85 89 99 Female Pneumonia 0 109 2 116 3 42 62 79 89 94 104 23 Now notice a third line that is new Study Designer ty Run All C Save as s X X Ev Y TIO admit X ee nea 67 Run Method Show Criteria a 5 Demographics Target Event All Events Patient List 3 m Sex Male Female Unknown n Male Pneumonia D baseline 741 100 0 0 0 0 P Female Pneumonia 0 0 905 100 0 0 et Minimum 5th 10th 25th 50th 75th 90th 95th 100th Male Pneumonia I line 0 098 1 207 2 232 22 57 75 85 89 99 Female Pneumonia 0 109 2 116 3 42 62 79 a9 g4 104 3 These are tabs within the TTO admit method canvas These tabs represent canvases within the TTO admit Canvas Each subcanvas conveniently provides outcomes relevant to the TTO admit method As you click on demographics Target Event All Events Patient List sequentially that tab becomes white and its associated sub canvas is filled with the relevant results of that tab Demographics tab provides the male female breakdown and age percentiles 24 ER Designer A Run All Savers VQ V 9X XC E 4 ok w TrO admit X Run Method Shaw Criteria Demographecs Target Event All Events Patent Lisi 3 Readmission for H F Preumonia Pts 1 2 B F T n p m uns mms 3 F b F 20 100 190 200 250 20 350 2 D
42. T TT 3 040 T warns ss races ee SSE EE Eee 0 ee a fs ep ct s ri ee ee sco Em en ce 0 50 100 150 Days pvalue 0 247 BadDiabetesWRepeatGoodF1 BadDiabetesWRepeatBa Cummulative with Event 30 Days 60 Days 930 Days 180 Days 365 Days BadDiabetesWRepeatGoodF1 3 8 6 7 8 3 16 1 baseline 1 8 5 7 4 1 9 2 5 5 11 1 12 3 19 7 BadDiabetesWRepeatBadri 7 10 37 So 136a 48 There is still no statistically significant difference between the two groups but we see something that looks paradoxical Those who achieve good diabetes control have a higher hospitalization rate We might hypothesize that those who were brought under good control suffered from more hypoglycemia and hypoglycemia associated hospitalization There is a short term cost in hospitalization in order to get the long term benefit Of course the difference is not statistically significant in the short term so we really should not be making too much of this This exercise has demonstrated the impact of the focus of the index event line which condition line the index event line points to on the nature of the question answered when the cohort is enrolled in the Time to Outcome Study The cohort index date carries within it the seeds of determining the question answered by the Time to Outcome Study in Study designer You must think carefully when you build your cohort anticipating the nature of
43. The first has the condition of inpatient admission Name Admit_0_ 365AfterDeadVvhenln ra Analysis Canvas c AD E Admit 0 365AfterDeadWhenIn Earliest of Admit 0 365AfterDeadWhenIn And 1 9577 Admit_0_365AfterDeadWhenIn All of Inpatient Admit InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date The second line has an event death that occurred during the admission during When In the event of the first line Name amp dmit Q0 365AfterDeadVhenln Analysis Canvas E AD Admi _0_365AfterDeadWhenIn Earliest of Admit_0_365AfterDeadWhenIn And hey Admit 0 365AfterDesdWhenIn All of Inpatient Admit InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date AND Eli DeathInHouseOrss All of DesthInHouseOrSS SSAorinHouse WHEN IN Admit_0_365AfterDeadWhenIn Inpatient Stay 75 Let s begin First use Approach 1 Go to Study Designer lt This opens up Study Designer CiG Study Designer ETTEN C amp s USR Admission Repeat Diabetic 5 ls USR Diabetes control outcome Left click on the S study for Usr Pneumonia Readmit Ear Pneumonia Readmit Left click on the pencil study edit CiG Study Designer ETTEN C USR Admission Repeat Diabetic USR Diabetes control outcome i USR Pneumonia 5 ls USR Pneumonia Readmit P S Moor x 76 USR Pneumonia Readmit Your study o
44. URISY W CC 194 000 P SIMPLE PNEUMONIA amp PLEURISY W CC 194 000 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 SIMPLE PNEUMONIA amp PLEURISY W CC 194 000 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 SIMPLE PNEUMONIA amp PLEURISY W MCC 193 000 RESPIRATORY INFECTIONS amp INFLAMMATIONS W MCC 17 SIMPLE PNEUMONIA amp PLEURISY W CC 194 000 mrem ON Ee r rns mo mte ntt n om 112 1 If you need identifiers and you have been granted privilege you will find a show identifier button which you can left click on to show identifiers 2 If you want to export the data Left click on the export button Export Method Export Type Excel2003 Export Excel2003 Excel2007 CSV Reset Close Choose Excel 2007 and then left click on export Export Method Export Type Excel2007 Export You can download your exported files here Data file Reset Close Now left click on Data File and the spreadsheet will open up with multiple tabs The first tabs are documentation for the cohorts and for the study method The last tab is the spreadsheet with the data 113 uo coc ic ae18ae23 4786 4be6 9634 b6732e1899e4 Read Only Excel z m Goon E HOME INSERT PAGELAYOUT FORMULAS DATA REVIEW VIEW ACROBAT
45. ame Pneumonia Readmit Optionally write a description Study Designer Watch what happens to the tab when you click Study Narfie Pneumonia readmit escription Enter Description below pm ui a 2 JO T Methods E 3 P Baseline Group Name Cohort Collection 4 1 1 e Group Name 1 No cohort collection selected 4 4 4 1 click here to select or build a group rrrrry Click on Groups tab Canvas 2 The study designer canvas switches from the Main Study designer Canvas to the Group Study Designer Canvas The Group tab turns white and the Main tab grays out d Study Designer Run All SaveAs gt OM X 1s Grp ID Baseline Group Name Cohort Collection 2 1 e Group Name 1 No cohort collection selected i x 23 5 i Ww 3 T o Al amp 4 39 4 4 i 4 2 note Groups tab is White indicating it is the active canvas layer of the study designer 4 If you want to build a new group cohort or collection you can click on the three dotted button This will open up the event canvas let you build your cohort or collection and then you can return to the study designer prefer to build my cohorts and collections first in the event canvas and then just select them from the drop down that becomes available when you Left click on No cohort collection selected 5 Hovering over No cohort collection selected 5 will activate th
46. ast Tutorial 24 List Method Long view https www youtube com watch vzhZegDjKyXzl amp listZPLf7raPnmlLOeAWU2cNf2jIMDRaQuUTC um amp index 24 Until now our ADs were singularities requesting the earliest or latest event Suppose you want to build an AD with All for example you are interested in All the Wbc in a patient with a pneumonia admission not just their first or latest Wbc To save time we will modify an old Analysis Definition and save it with a new name First add a new study method by left clicking on the sign CiG Study Designer X Study Designer J Run All LJ Save as y W 3X X s g USR Admission Repeat Diabetic sims V TTO Admit X Adv TTO Admit TTO Adv When In BasicWbcEcholist X s s USR Diabetes control outcome ______ sE sss E _ Reset Method Run Method Show Criteria s g USR Pneumonia Readmit List Pivot d Deom 5 _ Show Identifiers ri 4 l Page n of 106 b m Total records 1683 Select List as the Method Type and name the method All Wbc s during New Method Dialog x Method Type List Add Caneel l Click Add 153 Study Designer b3 Run All L Saveas y V PM X wy ES TTO Admit Lv Adv TTO Admit Y TTO Adv When In M BasicWbcEchoList All Wbc s During X Reset Method Run Method Main Groups No outcome selected Methods Method Name All Wbc s During Method
47. at phrase causing a box to appear around it and a downward arrow to appear on the right Left click on the downward arrow and you will see all the cohorts and event collections you have built available for selection Study Designer Run All O Save As ox X io Grp ID Baseline Group Name Cohort Collection 1 e Group Name 1 No cohort collection selected X x N o m O u D P 7 q q qd 4 4 f gt gt 10 study Designer 4 RunAM C saveas h V X X a Grp ID Baseline Group Name Cohort Collection M 1 e Group Name 1 No cohort collection selected v j x T E VJI vuwputeci ILVIDOMAZIITUILCIUI I mat ys LJ E 5 USR Pneumonia 2010 Female admits dc alive E USR Pneumonia 2010 Female discharge dc alive E 3 USR Pneumonia 2010 Male admits dc alive E 5 USR Pneumonia 2010 Male discharge dc alive E gt z USR Unique patients inpatient outpt 2013 E IR 4 USR Unique patients inpatient outpt or ED 2013 E gt C HEN Y C47 amp AT RA Ane arr Ft rw Use the vertical slider until you find the pneumonia cohorts you built before We want male and female pneumonia patients with primary diagnosis of pneumonia discharged in 2010 alive not expired Highlight the desired cohort and then put in a group name that will be used by the program to refer to this cohort or event collection in its graphs or summary reports St
48. ays pealue Q667 4 Male Pneumonia Female Pneumonia 5 The Target Event provides a graph that shows the name you chose elapsed time from the index date Y axis cumulative incidence of at least one admission by the associated date on the x axis Cumulative percent with at least one admission by that time p value of the comparison in this case not significant significance being 05 the legend for the two populations A vertical window slider bar so we can see the table below the graph 25 Study Designer Run All SaveAs e W X X TIO admit X Run Method Show Criteria Main Demographics Target Event All Events Patient List di 8096 70 t 60 o E 5096 o 4096 2 is 30 z E 2096 10 096 Groups m b Methods 0 50 100 150 200 250 300 350 Days pvalue 0 667 Male Pneumonia Female Pneumonia NO difference in readmission Cummulative 9o with Event 30 Days 60 Days 90Days 180Days 365 Days rate by Male Pneumonia 16 3 22 7 27 5 35 6 46 0 gender baseline 13 6 19 0 19 6 25 7 24 2 30 7 32 0 39 0 42 2 49 6 p 16 0 22 2 26 6 36 4 47 4 Female Pneumonia necks hep su xd 229 90 1 1440 hE 4696 of male pneumonia patients are admitted at least once by 365 days while 47 496 of Female pneumonia patients are readmitted at least once by 365 days Under these numbers are parentheses that provid
49. be com watch v wOvNybshJpc 4 Tutorial 20 A Reminder in Time ppt 7 https www youtube com watch v s77M_ mBdNnc 5 Tutorial 21 A conversation between objects ppt8 https www youtube com watch v lIfAZmzVrmSE 106 List Method Looking only at the attributes of the cohort Tutorial 22 List method listing attributes of the cohort https www vyoutube com watch Pv M99VtazqiIMo amp list PLf7raPnmlILOeAWU2cNf2jIMDRqQuU TCum amp index 22 equivalent to Browse in the Event Canvas with the added advantage that the results can be saved from session to session First we will do a quick and dirty method of just looking at attributes of the pneumonia cohort itself Now move to the pneumonia Readmit Study that you built before and Add a new Method the List method New Method Dialog Add Method Type List New Wubi Method Name SERERE e oT Add Cancel Creating a Method called Basic List for Cohort Left click on Add 107 Study Designer P Run All SaveAs V gt X X J N i i BasicListForCohort X 4 TTO admit Adv TTO admit asicListForCoho Reset Method JI Run Method u a B o Start End 3 Blackout 0 Days No outcome Time from Start 0 o I selected 3 1 Time Units Days lt p B To Advanced 1 Method Name BasicListForCohort 4 Method Description jst 4 1 gt gt 5 3 View Wide Long m Avai
50. build the study and the results of method of TTO admit 2 Left click on Show Criteria and you will find the criteria view and the button you just depressed changes its name to Show Results letting you know you can switch back to results by Left clicking d Run All L SaveAs g X 3X Study Designer E a TTO admit X Run Method un Q o 5 Start j End ui Blackout 0 Days Time from Start 365 t n q 1 ime Units Days 1 b To Advanced 4 1 Method Name TTO admit 4 Method Description Time To Outcome d d b Estimation Points 30 60 90 180 365 Days Graph Name Readmission for M F Pneumonia Pts You are now in the racetrack paradigm on the Method Canvas Note at the bottom of the racetrack is the phrase To Advanced 54 Hover over To Advanced and the button becomes yellow showing that it can be pressed left click Start End Blackout 0 Days Time from Start 365 Admission Left click on To Advanced and a new image appears DIT Analysis definition Start Blackout 0 Days o F B To Simple You are now in the Advanced Mode At the bottom of the Advanced Mode racetrack metaphor is the toggle button that brings us back to the Simple mode Toggle back and forth between the simple mode and the advanced mode Go on do a few more times Have a blast The Advanced mod
51. cessusecessuaeceeteneeetsgneses 153 Mame E 168 Pre class Assignments Read Pages 3 8 Build Cohorts on P 7 1 Tutorial 11 Temporal Relativity in Cohort Builds ppt3 https www youtube com watch v LHK9uLvDyoU 2 Tutorial 12 The cohort as an Analytic Object ppt 4 https www youtube com watch vzi2jCBlcPQuO 3 Tutorial 13 Introduction to Study Designer ppt 5 https www youtube com watch v qvPZonYO5AU Study Designer Event Canvas allows you to build a cohort or an event collection Both Cohorts and Event Collections are called Groups For ease of presentation we will use the word Groups rather than cohorts or Event Collections Once you have built your groups cohorts or event collection you will want to perform some analysis on the members of the group The collection of analysis rules applied to the group members is called a method When you have both a group and a method you have a study CLG study components CLG study group Analysis Method You notice the I on the image for the group This is to remind you that each member of the group has an index date time from which the analytic method measures elapsed time You learned about the within temporal operator on event canvas and you will be able to build in Study Designer Analytic Methods that will measure outcomes within a certain amount of elapsed time from the index date from each individual member
52. d Attributes Delete this x ale a attribute E 4 G 0 Common Attributes E o 4 2 0 4 g A 4 d amp G 0 2 EventName 4 o 1 4 i 3 p 3 SES ta N t G 0 4 EventID D 4 Apartment P T uos Deme Haase eee w lt E G 0 6 Start EventDateTime 3 D 6 State Q G 0 7 APR DRG Code D 7 CensusTract 6 0 8 Inpat Account D D 8 CensusBlock G 0 9 DualEligible 3 D 9 Latitude G 0 10 Admit Referral Agency O D 10 Longitude G 0 11 Discharge Referral Agency C D 11 CensusBlockGroup w v Restore Default 4 Study Status 4 Let s remove all the selected attributes except drg and disposition by holding down the keyboard control key and then left clicking on each attribute to be removed You will notice that all of your selected attributes are highlighted in yellow Then left click on the red x 111 Left click on Run Method Study Designer P Run AII save as Li tes x eS TTO admit u a o r E o y rrYrY The Result screen then opens up and you see a grid Study Designer Reset Method Run Method K View Wide Long u Available Attributes Outcomes Demographics Group G 0 10 G 0 11 G 0 12 G 0 13 G 0 14 G 0 15 G 0 16 G 0 17 amp G 0 18 G 0 19 G 0 20 G 0 21 Admit Referral Agency Discharge Refe
53. d Close NOT Disposition l Update and close to populate the definition condition https clg emerginghealthit com CLGNET Modules EventDefinitionBuilder EventDefinitionBuilder aspx DefType EV Name Inpatient amp dmitWDeath Event Type Inpatient Admit End Update Update and Close z Event Definition Options Luffy Disposition Equal EXPIRED Update and close again to populate the event definition In the GUI the entry box after points to the cohort collection date 84 Bl Edit Selected Edit Selected C Save As aaa OO Name Admit_0_365dAfterDeath I Analysis Canvas ECAD M New Analysis 3ef79cb All of Any And Lee Eventi New Event Definition within 0 to 0 select time units select time direction New Abstract Event Definition A Edit Selected Condition Admit 0 365dAfterDeath l1NOT Alof InpatientAdmitWDeath InpatientAdmissionDate WITHIN 0 TO 365 Days After Cohort Collection Date Update Update and Close Update and Close Edit Selected Name Admit Q0 365dAfterDeath i Analysis Canvas E AD New Analysis 3ef79cb All of Any And LS Admit 0 365dAfterDeath All of InpatientAdmitWDeath InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date The index event line still needs to be edited Remember that we are building this Analysis Definition for the purpose of using it in a Time to Outcome stu
54. dDiabetesWRepeatGoodF1 and BadDiabetesWRepeatBadF1 Choose from the drop down in study designer BadDiabetesWRepeatGoodF1 Studies GO d Study Designer J Run All SaveAs V X x S s USR Diabetes control outcome c i m s s USR Pneumonia Grp ID Baseline Group Name Cohort Collection s s USR Pneumonia Readmit 1 1 e Group Name 1 No cohort collection selected E X e C USR BadDiabetesWRepeatGoodF2 E c usr CHF Primary March 2012 Discharge E T c usr Inpatient Admit w CHF in Mar 2012 E 3 c usr Pneumonia Females 2010 Admit Primary E s c usn Pneumonia Females 2010 Discharge Primary E c usn Pneumonia Males 2010 Admit Primary E c usn Pneumonia Males 2010 Discharge Primary E C usR Bad DiabeteswRepeat E c usn Bad DiabetesWRepeatBadF1 E E Bad DiabetesWRepeatBadF2 E VY Ahhh Add another group choosing BadDiabetesWRepeatGoodF1 from the drop down 45 Name both groups using their drop down name Resulting in Study Designer A Run All C save as e pl X i Grp ID Baseline Group Name Cohort Collection 1 amp BadDiabetesWRepeatGoodF1 C USR BadDiabetesWRepeatGoodF1 E X L SU 2 2 BadDiabetesWRepeatBadF1 c usR Bad Diabetes WRepeatBadF1 e x in Uu E e E d Qu Add a Time to Outcome Method if you forgot how look at previous pages c Admit x 5
55. dy A Time to Outcome analysis will only recognize a singularity Earliest of or Latest of The default setting is a multiplicity All of Change the Index Event line to Earliest of the condition Admit O 365dAfterDeat h Right click the Index Event line hover over Edit and left click 85 HEER seve as EEr ha Name amp dmit Q0 365dAfterDeath InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date From the drop down menus choose Earliest of and then choose your event condition line AdmissionsAIl Edit Selected Condition oe Root Operator AND AnalysisDefinition Earliestof EARTE Update Update and Close Update and close AEs T saveas a29xl x Admit 0 365dAfterDeath Analysis Canvas E AD M Earliest of Admit_0_365dAfterDeath And Lere Admit 0 365dAfterDeath All of InpatientAdmitWDeath InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date Save and close CiG Analysis Canvas ATES Lal GBI Manage Analysis 3l Svs B system Analysis AD Py earliest gastro outpt vist AD Jj time to gastro i user User Analysis AD B usr Admit 0 36SdAfterDeath LOTI Admit_0_365dAfterDesth All of InpatientAdmitWDesth InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date 86 Your n
56. e Index Event ID Index Event Type Outcome E 13177 cdg 13177 1 365 0 0 365 2 26 2010 3 10 00 PM 860122 InpatientDischargeDate 15511 clg 15511 1 365 0 0 365 10 3 2010 10 37 00 AM 916952 InpatientDischargeDate 23278 cig 23278 1 2 1 2 31 3 23 2010 4 54 00 PM 866017 InpatientDischargeDate 3 25 2010 23839 clg 23839 1 22 1 5 99 10 26 2010 3 10 00 PM 921403 InpatientDischargeDate 11 18 201 24380 cig 24380 1 365 0 0 365 3 2 2010 5 37 00 PM 861409 InpatientDischargeDate 24794 clg 24794 1 365 0 0 365 1 27 2010 1 03 00 PM 847781 InpatientDischargeDate 32541 cdg 32541 1 11 1 1 62 1 30 2010 7 15 00 PM 850488 InpatientDischargeDate 2 10 2010 40330 clg 40330 1 139 1 1 365 6 22 2010 3 11 00 PM 889848 InpatientDischargeDate 11 8 2010 40826 clg 40826 1 319 1 1 365 12 8 2010 9 07 00 PM 933294 InpatientDischargeDate 10 23 201 48595 clg 48595 1 54 1 1 365 4 29 2010 4 24 00 PM 874350 InpatientDischargeDate 6 22 2010 50620 cdg 50620 1 5 0 0 5 9 8 2010 2 13 00 PM 905802 InpatientDischargeDate 51210 cdg 51210 1 365 0 0 365 2 22 2010 4 27 00 PM 858751 InpatientDischargeDate 51403 clg 51403 1 365 0 0 365 10 25 2010 6 45 00 PM 922352 InpatientDischargeDate 51720 clg 51720 1 244 0 0 244 2 13 2010 2 17 00 PM 856821 InpatientDischargeDate 54080 cig 54080 1 365 0 0 365 4 10 2010 2 40 00 AM 871963 InpatientDischargeDate 55219 clg 55219 1 365 0 0 365 2 15 2010 8 50 00 PM 855838 InpatientDischargeDate A restricted data set This is not considered anonymized because it has dates of ser
57. e l7 ic WALNUT TREE WALNUT TREE Event Definition Options 1G Study Designer VOLUME TRANSFUSED Update and Close WALNUT TREE CLASS WALNUTS WALNUTS CLASS WASHED PLT ASSAY WATERMELON WATERMELON CLASS WAVE FORM WAXY CAST WB AGG INTERP a Definition i ESF undefined Condition IB Aggregation WBC NOT Lab Test Type y WBC by EN CEN Update and Close m BIK Please Select EVENT amp SelMenuOptic 131 Definition Canvas Webpage Dialo J e https clg emerginghealthit com CLGNET Maodules EventDefinitionBuilder EventDefinitionBuilder aspx 7Def l ype EV I Edit Options D Name WBC Event Type Lab Test Update Update and Close E Event Definition Options l L y Lab Test Type Equal WEC Update and Close We now see the new Event Definition populated with the WBC LabTestDate event we just built F Edit Selected Condition First WBC NOT Earliest of WBC LabTestDate Y WHEN IN Update We want to force the White Blood Count to occur during the hospitalization We want the White Blood Count When In the cohort index event Left click on New Duration Definition and choose Cohort Collection Duration By doing so White Blood Count will be sought outside of the analysis canvas and during the cohort members index event the original hospital discharge New Duration Def Select from
58. e paradigm is similar to the simple mode Like the Simple Mode the Advanced Mode has a start flag The Advanced Mode has an end flag of sorts but its end flag is HUGE 55 Unlike the Simple Mode the Advanced Mode does not have an Outcome Event Box Somehow in this HUGE End Flag contains information for both the Outcome Event and the notion of elapsed time This will give us tremendous flexibility This Huge End Flag is called an Analysis Definition The Analysis Definition contains information about elapsed time and whether the event is Earliest or Latest IMPORTANT POINT Simple Mode Assumptions Made Explicit The simple mode is simple to enter but makes certain assumptions Time is measured in the forward direction from the index date until the first downstream event The most proximal event in the forward direction determines whether an event has happened or not When the direction of the racetrack is backwards when we look prior to the index date to find an event by pressing the F in the racetrack image the F becomes a B Cr Blackout 0 Days Time Units Days To Ad ced lt B End Time from Start 365 The event that is now sought is the first event going backwards in time from the index date This event is the most proximal event in the backward direction from the index date From the perspective of Time s arrow starting at 365 days before the index date and then ending at the index date the event mo
59. e the 9596 confidence interval for the observation You notice vertical lines in the graph These are censorship lines They are recorded every time an individual patient is censored The individual can be censored by death or by run out of time Since all the patients were seen in 2010 all of them had the ability to be seen for a year unless death censored them Had the study involved patients admitted only two months ago then those patients would have been censored at month two and we would have seen a large number of vertical lines We now click on the All Events tab 26 Study Designer d Run All j saveas X X gv TITO admit X Run Method l Show Criteria a 3 Demographics Target Event All Events Patient List 5 wn Number Target 9 i Events Achieved Person Days at Risk Incidence Density per year of risk Relative Risk Risk Difference 5 Male Pneumonia 1 215 baseline iii co 1 132 1 303 MN uu Female Pneumonia 1100 298194 nal age siis is i Note Relative Risk group baseline group Relative Difference group baseline group In this tab we count all the downstream admissions in the next 365 days not merely the first We also count the number of person days of observation in each group Unless a person is censored by either death or administrative run out of time each patient will contribute 365 days of follow up time We then calculate the third column inc
60. el 78 Rename the Method to Advanced TTO Admit J New Method Dialog Method Type Time to outcome 43 8 Method Name Advanced TTO Admit Click Add and your new method will appear as a tab in study designer SNPs Run All O save as O GX TT l 0 n VW Adv TTO Admit X T Main End Time from Start 0 S Start Blackout 0 Days No outcome ed Methods Groups Time Units Days B To Advanced ED Now click on To Advanced You see the Advanced Method with the Analysis Definition Flag replacing the End flag 79 Study Designer Jv Run all _ Save As z 9X x Doo i it X s lt gt TTO admit Adv TTO Admit u 3 ini Ty Analysis definition e M Ivsis definit 5 Stat TERM No analysis definition selected 7 4 u Blackout 0 Days 3S J o o Time Units Days BD To Simple You now must build an Analysis Definition To build left click on the three dotted button Click on the builder button and you will find yourself in the Analysis Canvas Analysis Canvas J Management gt m EES e Manage Analysis Name New Analysis 8dba9f7 di a sys System Analysis sss earliest gastro outpt visit TA i eee o Wt New Analysis 8dba9f7 All of Any And User user Analysis New Analysis 8dba9f7 Ap B USR adl around 3d I Cam B uco obo irct A 265 roe Our goal is to
61. erD Copy from z FINOT Alot v New Event Def v WITHIN 0 TO 0 Days After Cohort Collection Date Update Update and Close The new Event Definition Palette will open g ver 7 fallen a iem Big P D ote iz amesgingheshh com CL Ghi Ahatia EowrtDefineteriducdar our wh ins radit aga Daf V ica z EANTA Sama Opt arc Hd cmmDatim ten rtm AD adorant Daf ita n gust Wal 7013 8225 3 08 aU 5 124006 ib EI Fvent arme ty Cted N Type EPE Pears Ti oct End Point Event Messe Select 82 Name the event definition InpatientAdmitWDeath and choose Inpatient Admit as the event type Now define your event attribute Right click on Definition and add a condition gl Definition Canvas Webpage Dialog g e https clg emerginghealthit com CLGMET Modules EventDefinitionBuilder Eventl Jmm Name InpatientAdmitWDeath Event Type Inpatient Admit Update Update and Close Event Definition Options E pr E Sub Group A new condition appears in the GUI editor on the bottom of the screen From the drop down menu select Disposition from the list of inpatient admission attributes u n Another drop down will appear to the right of disposition You can select or create a set u n Select Then select expired from the last drop down on the GUI editor aF Ta Tea Ta E Definition
62. esigner Study designer opens up CG i Study Designer Now left click on the sign to create a new study Study Designer tk Run All LJ save As lg Lime X m g c 1 r m Study Name New Study c9fc 4 un Description m 3 5 228 4 k m 5 Methods 3 8 P Baseline Group Name Cohort Collection 4z 1 1 Group Name 1 No cohort collection selected 6 4 4 1 gt E On the left hand side of the study designer canvas you see three tabs Each tab represents a canvas within this particular study 1 Main this is the Main Canvas of this study where you name and save the study The Main Canvas is a summary canvas for the study It will ultimately include the name of the groups on the left hand side and the name of all the methods you have built in this study to act on those groups The methods will be listed on the right hand side of the page 2 Groups This canvas has the names of each group you are including in the study These are the groups upon which the study methods are applied A group can either be a cohort or an Event Collection Each group member is anchored at his index date from which elapsed time is measured 3 Method This is the Method Canvas You are allowed to build multiple methods on this canvas as will be demonstrated These methods are composed of three types Time to Outcome List or Time in Range Exercise 1 Pneumonia Readmit Study Give the study a n
63. ethod Type New DEVE TATT Method Name Select a Method Type Time to outcome N List Time in Range Add Cancel 13 sulle Method Name TTO admit Left click on Add Notice at the top line next to the 4 sign a new tab has appeared This tab is the TTO admit outcome canvas You are able now to instruct study designer how you want the TTO admit method to function 14 Study Designer Fy RunAl L Save As aX o 2 a Groups Methods M ee TTO admit X 1 Run Method Start End Blackout 0 Days No outcome o4 from start 0 selected Time Units Days B To Advanced Method Name TTO admit ini Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days 8 Graph Name KM Graph Name 9 will now give you a tour of the TTO admit Method canvas Numbers are keyed to the image above 1 Note how the tab telling you are on the TTO admit canvas is white When you have many methods they will be added next to this tab but will be grayed out if not active When active you can modify the criteria However until you run the study the results do not change Note the method tab is also highlighted White This is a visual cue that you are on a method canvas not a group not a main The top row TTO admit tab tells you that within the method canvas you are in the TTO admit canvas These two tabs the vertical met
64. ewly build Analysis Definition will appear in the management pane CiG Study Designer Analysis Canvas E Management Pinas USBI Manage Analysis EC eariiet getro oulpt vit P LAD P time to gastro E la user user Analysis USR Admit 0 3e5dAfter USR Admit 0 365dAfterDeadWhenIn Left click the refresh and close button and you will be directed back to Study Designer Find the new Analysis Definition from the drop down menu Study Designer de Run All _ Save As x x py 4 Ea TTO Admit y Adv TTO Admit X Main Run Method ll Show Results Mn iA Analysis definition JSR Admit OQ 365dAfterDeath Groups No analysis definition selected kh d E _ f USR AdmissionFirst365DeathEarliest z 4 F USR Admit 0 365AfterDeadWheniIn SG no secon B ore F 0 Method Name Adv TTO Admit F AdmitO 365dAfterDeathAttribute Method Description Time To Qutcome F USR FirstAdmission0_365days FE USR Inpatient Admit E InpatientWAtributeDeath Estimation Points 30 60 90 180 365 Days Graph Name Adv TTO Admit Choose the Analysis Definition Admit O 365dAfterDeath and Run Method Study Designer P Run All C Save as e iX X E TTO Admit ied Adv TTO Admit X Show Results ania ta Analysis definition Start Blackout 0 Days un a r A Qu o To Simple Method Name Adv TTO
65. f 3 D 5 city a 5 G 0 16 Discharge Attending 3 p 6 State Q 8 G 0 17 Discharge house staff 3 D 7 CensusTract G 0 18 Facility D 8 CensusBlock G 0 19 Insurance Plan I D 9 Latitude G 0 20 IPA Insured 3 p 10 Longitude G 0 21 Line of Business 3 D 11 CensusBlockGroup Restore Default Now you see DRG in the Selected Attribute List Selected Attributes D 16 D 17 D 18 D 19 D 20 D 21 D 22 D 23 Patient Name Primary Care MD Set Patient Phone Preferred Language Race Zip Code Set Age at Index Date Social Security Number Present G 0 3 EventDateTime G 0 38 Disposition G 0 39 Inpatient ICD9 Group G 0 12 DRG M View Wide Long 4 2 Demographic attributes can be made visible when you left click on Demographics tab 2 110 3 Later we will show you how to build an outcome using the simple method racetrack paradigm or advanced method using the Analysis Definition to pull the outcome attributes over to selected outcomes as well If you want to eliminate a previously selected attribute you have to left click on it and then left click on the red x upper right hand side JEM peers y Run All O savers 9x x S me TTO admit W Adv TTO admit BasicListForCohort X z i Reset Method Jl Run Method a 2 View Wide Long m 3 OQ Available Attributes Selecte
66. finition After choosing the Echo we note the following kc m m 1m 1m 4 Witte Analysis definition No analysis definition selected USR USR WSR USR USR Method Name VSR CardiacEchaEF180DaysBefore Admit 0 365dAfter Admit OQ 365dAfterDeadWhenIn Admit O0 365dAfterDeath Inpatient Admit WbcWhenInFirstCohortDuration Method Description List Notice how the Outcome End flag Analysis Definition moved to the left This is a result of the fact that the AD is before the index date time 148 int Analysis definition Start Blackout 0 Days Maximize the list by left clicking on List populator Maximizer Available Attributes P 0 0 Common Event Attributes gt UU 0 1 WbcFirstWhenInCohortDuration X 4 0 2 CardiacEchoEF180daysBef X 4 CX 0 2 1 Ejection Fraction 0 2 1 1 Count 0 2 1 2 EventID 0 2 1 3 EventType E 0 2 1 4 EventDateTime 0 2 1 5 EventName 0 2 1 6 Aortic Area 0 2 1 7 Aortic Mean Gradient 0 2 1 8 Aortic Peak Gradient L E Move the window slider down until you see Best Value Ejection Fraction 149 Available Attributes 0 2 1 13 Aortic Valvular Calcification 0 2 1 14 Aortic Valve Description 0 2 1 15 EchoComments 0 2 1 16 Echo Conclusions 0 2 1 17 Echo Procedure 0 2 1 18 Echo Room 0 2 1 19 EF Alt Best Value 0 2 1 20 EF Best Value
67. he first subsequent admission The analysis definition is measuring elapsed time from the cohort discharge date until the first admit date time You might wonder why we bother with an advanced method if we can obtain the same results in the simple method There are things you can do with the Advanced Method that are impossible with the simple method You can do a time to outcome to the latest not just the first You can build multiple condition lines in the Analysis Definition impossible in simple method In list method next lecture you can use All of not just Earliest or Latest In list method next lecture you can use a when in duration to list things that occur during the hospitalization op i qe Now let s proceed to show why the Advanced Mode is important Let s create an Analysis Definition where there are two condition lines something impossible in the simple method 73 Complex TTO Examples requiring an Advanced TTO Tutorial 18 Time to Outcome Example requiring Advanced Method https www youtube com watch v VChMwnymKCw amp index 18 amp list PLf7raPnmlLOeAWU2cNf 2jIMDRqQuUTCum Suppose we wanted to identify elapsed time from index until the first admission in which the patient died The target admission will not be the first admission if the patient did not die in the first admission If the patient dies in the second admission then the elapsed time goes from the index date of the cohort u
68. he other 99 would be blank Creating such wide spreadsheets stresses CLG a great deal so when you have all as your target outcome you should use Long View instead of wide view It also makes the data unmanageable III View Wide Long s In the long view there will be individual rows for each wbc for each patient 165 The patient with 100 wbc s would have 100 rows repeating his medical record number but each row would have a different wbc value Drag Lab Test Value and Count from Available Attributes to Selected Attributes Reset Method Run Method Outcome Analysis definition Start ahs USR WbcWheninAllCohortDuration Y Jad Blackout 0 Days U Time Units Days Be To Simple Method Name All Wbc s during Method Description List Available Attributes 0 1 1 7 Grouped Numeric Labs 0 1 1 8 Lab Text Dont H 5 O 1 1 11 Test MD 0 1 1 12 True First omes Demographics Group gt O 1 1 13 Lab T View Wide Long gg Selected Attributes 113 D 1 PatientID D 2 MRN amp G 0 3 EventDateTime isposition G 0 41 Inpatient ICD9 Group 0 1 1 4 EventDateTime O 1 1 13 Lab Test Type E Both Count and Lab Test Value are in Selected Attributes Available Attributes Agr Une CVELE UE LII 0 0 3 EventID e 0 0 4 EventType 4 UW 0 1 WbcWhenInAllCohortDuration X 4 OX 0 AII WBC P 0 1 1 1 Count 0 1 1 2 EventI
69. hods and horizontal TTO admit orients you to where you are in the study designer Take a few moments to reinforce this in your mind Some students click through this quickly and then find they have lost their orientation By spending a few minutes with this paragraph you will become comfortable knowing where you are in study designer Paradigm 3 4 5 The paradigm we are showing you for data entry in study designer is the racetrack Every patient in each cohort has his own index date so each one has his own start flag beginning on that index date 15 We are in what is called the simple mode of study designer We follow that patient forward in time note the F on the racetrack to the end of the race the end of time specified in the stop flag 4 If you want to go backwards in time you left click on the F button on the racetrack and the flags reverse and the racetrack has a big B instead of F Outcome Start End Time from Start 0 No outcome Blackout 0 Days selected Time Units Days To Advanced The outcome 5 we are looking for is placed in the outcome image 5 Let s review each image on the racetrack and input the relevant data for our time to outcome admission analysis First left click on the start flag 1 N End Blackout 0 Days No outcome E from Start 0 selected gt B To Advanced Time Units Days The start flag control opens up 0
70. idence density per year of risk Of note now Female pneumonia patients have a higher incidence density When we divide the Female by the Male incidence density we get a relative risk greater than 1 1 108 and a 95 confidence interval that excludes one which means the results are statistically significant This suggests that when you consider all admissions not just the first women have a higher readmission rate If you forget what is being divided by what look down the canvas and read the Note The last column is the risk difference model Risk difference calculates the group risk the baseline group risk In our case this is the Female risk male Risk We see the difference is positive and zero is excluded so this is in fact significant Women have a higher risk by the risk difference method when you count all admissions downstream not just the first Warning Before you draw any conclusions you might want to look at mortality rate downstream of the first readmission as the Dead do not get readmitted Left clicking on Patient List tab results in 27 Study Designer Run All SaveAs p W gt X X QUY V TrOadmit X el Run Method JL Show Criteria A c Demographics Target Event All Events Patient List Ust Pivot 3 _ Show Identifiers 1 2 Export E M4 4 Page 1 of 103 b ppl Total records 1646 PatientID MRN GROUPNO TimeTo Occurred Total AllRisk Index Dat
71. in ne IMI geras neon sg usc canon Shy Kare Pneumonia Besckrst Darien s fiusa cancer Ped 8 1 5 gl US Compare tica up beo years Baseline shame 3 TTO admit Cohort Collection ma a usse menos corel outcome L Hias Drei c ui veteran 20 LO Hale dicha de aie E Tema To Outcome s iniinis 21 era Pria sanari cusa Prepares Ob Forrade d charge dc ale E 5 jg MPSA Cenc cipt Folkow un 3 ji usn vnd s paree somi pelbis CABG Nob fS Note you are opened directly to the Main canvas of this study showing you the study objects 1 On the left are the groups 2 Ontheright the Methods applied to the groups In this case there is only one method Left Click on the Method tab ESSIGPAPESIUIEE Run All O savers fe i gt 3 c 4 TTO admit X Run Method Show Criteria e Demographics Target Event All Events Patient List mo 8 S Ip 741 100 0 0 0 0 Female Pneumonia 0 0 905 100 0 0 o g Minimum Sth 10th 25th 50th 75th 90th 95th 100th Male Pneumonia T E z baseline 0 098 1 207 2 232 22 57 75 85 89 99 Female Pneumonia 0 109 2 116 3 42 62 79 89 94 104 53 1 The Method tab becomes white and the other tabs gray out 2 We note that the method canvas is open and displaying results in the results mode That is why the 2 Show Criteria button is offered You can easily toggle between the criteria used to
72. ion Hover over To Advanced and left click to move to the Advanced mode Outcome End Na pris Time from Start 0 JE S Time Units Days To Advensed Choose from the Analysis Definition dropdown No analysis definition selected and left click on the triple dotted button 156 Outcome Analysis definition i F Time Units Days To Simple Method Name All Wbc s during Method Description Liszt An Analysis Definition skeleton appears on the Analysis Canvas We are going to exit the Analysis Definition canvas go to management pane of the Analysis canvas remove the skeleton edit a previously built AD and save the newly created AD with a new name Click on the X Analysis Canvas lin v AERE T saveas EEX 4 Name New Analysis 78dd411 C System Analysis earliest gastro outpt visit r Analysis Canvas k AD Bj time to gastro AD f New Analysis 78dd411 All of Any And New Analysis 78dd411 USR Admit 0 355dAfterDesdWhenIn USR Admit 0 365dAfterDeath USR CardiscEchoEFI80DaysBefore USR WbeWhenInFirstCohortDuration 157 The canvas now looks like this Analysis Canvas T Svs B system Analysis LAD EJ earliest gastro outpt visit CAD B time to gastro User user analysis AD New Analysis 7add411 CAD B USR Admit 0 365dAfterDescWhenIn j AD usn Admit 0 365dAfterDeath AD Bj usr Inpatient Admit USR WheWhenInFirstC
73. is Canvas AD New Analysis 7787263 All of Any And We are interested in the Cardiac Echo Ejection Fraction s that occurred before the inpatient discharge We require the cardiac echo to have occurred within 0 180 days prior to the cohort index date Therefore we need a Within condition Right click on New Analysis and Add Within 142 A new condition line will appear We need to find the latest Echo for each patient 0 180 before the index discharge In the Gui editor below change All of to Latest of Specify the timeframe 0 180 days Before the Cohort Collection Date ECAD ij New Analysis 7787263 All of Any And LEE Eventi New Event Definition within 0 to 0 select time units select time direction New Abstract Event Definition Edit Selected Condition Ejection Fraction MOT Latestof New Event Def WITHIN 0 TO 180 Days Before CohortCollection Date Create a new event definition for the Cardiac Echo Hover over New Event Definition and left click 143 New Event Def dm meg SEEC from See Select from Common Defintion gt Ejection Fraction Copy from SNOT Aor New Eneo WITHIN 0 TO 180 Days Before C ENCTTENN Cy NNCOTTAANN A blank definition palette appears E Event Definition Palette Webpage Dialog e https clg emerginghealthit com CLGNET Modules EventDefinitionBuilder EventDefinitionBuilder aspx DefTy
74. l show you the options you have for units of time 32 Outcome Start End Blackout 0 Days Time from Start 365 Admission B Months 10d Name TTO admit Years 10d Description Time To Outcome Hours Minutes Seconds nation Points 30 60 90 180 365 Da As a rule of thumb if the interval O to 1 is important to you you should choose a unit of time measure smaller than the unit you are interested in 33 Challenge Problem 1 Tutorial 15 Challenge Problem Hospitalization rate post attempt at Diabetes Control https www youtube com watch v GprwqUZsQ A amp listZPLf7raPnmlLOeAWU2cNf2jIMDRqQuU TCum amp index 15 Using Cohorts you have built and the new skills you have learned solve the following Problem You want to know whether Diabetic Patients who initially have horrible HgbA1c greater than or equal to 9 5 who within 180 days to a year are brought under good control repeat HgbA1c less than or equal to 7 have a higher or lower hospital admission rate after control compared with Patients with a horrible HgbA1c whose HgbAic on repeat study six months to a year later are greater than or equal to 9 First build a mental map of the question Diabetes example temporal map 1 Build Cohorts 2 Study Outcome Outcome o Fimeto Event gg 180 lime to Event 365 Time to Event 365 New TEM Pobre a s e emule Sra t Cohort 1 Hospitalization Aic Atc lt 7 0 Index A1C 7 Rate
75. lable Attributes Selected Attributes 1 N ES 4 G 0 Common Attributes 7 J G 0 2 EventName Restore Default 4 Study Status a IN v v phics Group To get information just on the cohort not on any outcome we can ignore the racetrack paradigm and go directly to the list populator below Note you are ignoring both the simple method and the advanced method because all you are doing is looking at the attributes of the cohort itself You are going immediately to the bottom half of the screen an area that am referring to as the list populator We note symbol 1 which is a toggle switch that will maximize the image of the List populator The list populator moves attributes from available attributes 2 to selected attributes 3 Some of the selected attributes are brought over by default Maximize the list populator section of the screen by Left clicking on 1 Resulting in 108 Study Designer a Run All saveas X X j i i BasicListForCohort X 4 gt TTO admit Adv TTO admit asicListForCoho Reset Method Jl Run Method a o View Wide Long m BIg Available Attributes 4 Selected Attributes 4 5 u a 313 4 G 0 Common Attributes a c 5 09 z 8 fc G 0 2 EventName 25 on c l p 3 sES 6 L a G 0 4 EventID D 4 Apartment gt G 0 5 EventType 3 s city 1 g i 3 E G 0 6 Start EventDateTime f
76. mary E Pneumonia Males 2010 Discharge Primary E TTT Ach i E 4 Name the group BadDiabetesWRepeatGoodF2 5 add another group 6 Nameit BadDiabetesWRepeatBadF2 Your Study Designer Canvas should look like this a Study Designer o Run All _ Save As ox x m Grp ID Baseline Group Name Cohort Collection n 1 amp BadDiabetesWRepeatGoodF2 C USR BadDiabetesWRepeatGoodF2 E X t Paii y 1 Ir j 1 c 2 gt BadDiabetesWRepeatBadF2 C USR Bad DiabetesWRepeatBadF2 E x 5 zd ut T a A dE Qi 7 Left click on the Method tab to get you to the Method Canvas 37 Study Designer L SaveAs X X 8 Leftclick on plus sign to Add a method 9 Choose Time to Outcome as your method 10 Name the method Admission Resulting in New gue Method Name Admission 0 0 0 0 0 0 Cancel 11 Left click on Add 38 Study Designer d Run All O save as a gt x X m E 4 gt Admission X Run Method u 3 Outcome o 5 Start End ui Blackout 0 Days No outcome Time from Start 0 F selected de ES Time Units Days IL B To Advanced Method Name Admission Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name KM Graph Name 12 Now fill in the Racetrack Metaphor in the simple
77. mode a Openthe end flag and put in 365 days Close End flag Open the outcome box c Choose from drop down admission You built this in your previous Pneumonia example d Close outcome box e Keep Method Name as is f Keep Estimation Points as is g Change the Graph Name to Hospital Admission Your Study Designer Method Canvas should look like this 39 Seale y Run All J Run All O save As aX X Admission X Start End Blackout 0 Days Time from Start 365 Admission B To Advanced Method Name Admission Method Description 3 Estimation Points 30 60 90 180 365 1 1 Days Graph Name Hospital Admissio 13 Run Method Study Designer Run All saveas k vo X X y Admission X Run Method Show Criteria Demographics Target Event All Events Patient List SEN n ce ida a 160 42 9 213 57 1 oo BadDiabetesWRepeatBadF2 375 37 13 635 62 87 0 0 ieee 21 36 39 45 52 58 62 64 65 BadDiabetesWRepeatBadF2 21 31 36 43 51 58 62 64 65 14 Save the study 40 15 Study Designer D Run All J Save As le lai b X x lt gt E Y Admission x Run Method Show Criteria Main Demographics Target Event All Events Patient List E A xus 5 baseline 160 42 9 213 57 1 0 0 BadDiabetesWRepeatBadF2 375 37 13 635 62 87 0 0 Methods Groups Age Percentiles AU Y ERE He I EI n
78. ntil the admit date of the second admission which is the admission in which the patient died Our analysis definition requires that we indicate in some way that a death was involved in the admission that followed We can express death during a hospital admission in two ways Approach 1 One approach is to appreciate that when a death occurs in a hospital admission that admission is profoundly changed by the occurrence of death The admission is subsequently branded by the death during the admission The attribute of death will then be permanently associated with the hospital admit event A hospital admission with the attribute death can be defined by choosing an inpatient admit event type and then adding a definitional condition Disposition Expired x Name Inpatient Admit oS CS NNSCOTKNNNN 2 Event Definition Options 74 Approach 2 The advanced mode uses an Analysis Definition permitting many condition lines to define the event of interest The simple mode only permitted a single event and not a series of related events The first condition line will be an inpatient admission without any condition of disposition expired Note the event definition in the image below There is no condition of disposition expired E Edit Options Name Inpatient Admit E Event Definition Options Update and Close You will build on the Analysis Canvas two condition lines
79. ode 1 3 D 15 Gender I Restore Default 154 Highlight all Selected Attributes and delete View Wide Long Selected Attributes k The only attributes that remain are those required by the system Study Designer P Run All O Save As o 3 x 7 Im TTO Admit Adv TTO Admit LY TTO Adv When In BasicWbcEchoList All w Reset Method Run Method View Wide gt Long m Available Attributes Selected Attributes 4 G 0 Common Attributes G 0 1 GroupNo G 0 1 GroupNo D D 1 PatientID Methods G 0 2 EventName D 2 MRN G 0 3 EventDateTime A G 0 4 EventID amp G 0 5 EventType G 0 6 AdmitToInpatFromED Order Datet J G 0 7 LengthOfStay Analytic days Outcomes Demographics Group A G 0 8 LengthOfStay days wwwwwwz A G 0 9 Admit Nursing Station G 0 10 Admit Priority G 0 11 Admit Referral Agency G 0 12 Inpatient Admit Service Group G 0 13 APR DRG G 0 14 APR DRG Average LOS G 0 15 APR DRG Code 155 Minimize the list populator by clicking on the list maximize toggle switch View Wide Long K 3 0 1 GroupNa 1 PatientID D 2 MEN The Simple method assumes that we are only looking for a single value To obtain All Wbc s we need to go to the Advanced method and build an Analysis Definition We will go to the Advanced method and modify a previously built Analysis Definit
80. of the group da id Once you build a study you will see how tightly bound the group and its analytic method is They can be bound through the index date of the group How much time has elapsed from the index date to some outcome Analysis Method Also they can be bound with a When IN duration operator If the Group Member is identified by a durational event like admission the Analysis Method can relate to the individual group member through the duration of the group itself For example you can find all the cardiac echoes Analysis Method cardiac echo When In the Group s admission event When you want to compare outcomes between groups you put more than one group in the study Male diabetics Female diabetics a Death in 1 year If you want to do additional analyses in a study you add more methods Male diabetics Death in 1 year Female diabetics 4 Readmission 1 year So a study can have more than one group and more than one method In this Self Study guide you will learn how to build studies composed of multiple groups and multiple methods Analytic Methods There are three analytic Methods available to you in CLG These Methods will provide you with most of what you will need to extract meaningful information In addition they will also allow you to export data so you can build more complex models in the statistical package of your choice The three Analytic Methods are provided below
81. ohortDuration Hover over the new analysis definition in our case New Analysis 78dd411 yours will be different and right click Delete the new analysis definition Analysis Canvas p Management ESL x api Manage Analysis Svs B System Analysis AD f tme to gastro s PAN AD BJ USR Admi o 3 158 Now it s gone Analysis Canvas We will modify WbcWhenlnFirstCohortDuration changing it from First to AI Right click on WbcWhenlnFirstCohortDuration and Edit Analysis Canvas Management Banas UAB Manage Analysis i Sys Bl system analysis time to gastro View Summary share 159 Resulting in Analysis Canvas ziom T saveAs X X Name WbcWhenlnFirsticohortDuratian analysis Canvas E AD JJ wbewherinFistCohortDuration All of Any And EE First WBC Earliest of WBC LabTestDate WHEN IN Cohort Collection Duration The first order of business is to prevent the overwriting of your previously built Analysis Definition To do this you will change the name and check off Save As Check off Save As and change the name to WbcWhenlnAllCohortDuration E a ai Name WocWhenlnAllCoh ration Analysis Canvas LAb B WbceWhenInFirstCohortDuration All of Any And Eit First WEC Earliest of WBC LabTestDate WHEN IN Cohort Collection Duration Right click on the Firs
82. on line of interest Event AdmissionsAIl Even though there is only one condition line it is good practice to always point to the line of your interest Resulting in Edit Selected Condition AnalysisDefinition Earliestof Event AdmissionsAll Root Operator AND Update Update and Close Now update and Close and the Analysis Canvas Looks like X P _ ARREST saveas ou Uo B Earliest of AdmissionsAlll And Se ee AdmissionsAll All of Inpt Admit Save and close and you will see your Analysis Definition in the Management Pane 65 p Analysis Canvas A Management o x 2 Rbi Manage Analysis m Svs System Analysis E earliest gastro outpt visit P time to gastro 4 user User Analysis g USR AdmissionFirst0_365days lt 1 u 1 Note the Saved Analysis Definition AD in the Analysis Canvas Management Pane 2 Now Exit and return to the study Advanced Method 66 MEEA Jy Run All O saeas D ex X 4 P 4 VA TIO admit f Adv TTO admit X ficco Analysis definition Start No analysis definition selected Blackout 0 Days B To Simple Method Name Adv TTO admit Method Description Time To Outcome 7 0 Estimation Points 30 60 90 180 365 Days Graph Name KMGraphName
83. ospitalization rate p 978 This is reassuring in that it suggests that the two cohorts are comparable at baseline and that what we have seen post therapy in hospitalization rates was the result of the control not to baseline differences But wait look closely at the graph above and the table at 180 days 47 By day 180 16 196 of those fated to be in good control were hospitalized at least once whereas 13 696 of those fated to be in bad control were hospitalized at once Surprisingly the good diabetics are hospitalized at a faster rate than the bad diabetics Let s see how strong this observation is Let s repeat the Time to Outcome analysis but modify the follow up time in the End Flag to 180 days This will let us compare the relative hospitalization rates in the two groups considering only the first 180 days of follow up time Time from Start 180 Days v Start _ Event Based End pater No analysis definition se ected Blackout 0 Days _ Uploaded End Dates p Time Units Days Be To Advanced Yielding Hostpitalization Rate 100 T T LLTLLLM E00 T ILMMM Meell d i 80S T T 670 E60 T T L M LIIMMe eMM ll l S acc aaa 40 T TT TMMeell0 T B30 TTL TMM ee e l 2 E 20 T
84. our cohorts are pointing to the second event 35 Here are the steps d Study Name EElSr Nev dig Rettore pr E m ui Description Enter Description oO z 9 4 o z 4 Baseline Group Name QU 1 Group Name 1 Mh le sl 2 Activate Group tab f Study Designer Grp ID Baseline 1 Group Name e Group Name 1 Groups of group is active tab as it is white same as the active canvas 1 Methods Jy Run All save as In Main tab Name the study Diabetes Control Outcome ETI 3 runai O soer ex x _ Methods Cohort Collection No cohort collection selected x x Cohort Collection No cohort collection selected m x 2 left click on downward arrow next to No cohort collection selected and choose the BadDiabetesWRepeatGoodF2 as first group Note the good controlled group is the baseline 36 1j Study Designer x Run All _ Save As H x T Grp ID Baseline Group Name Cohort Collection a 1 ic Group Name 1 zd x i21 i j Mo cohort collection selected o b c usn Bad Diabetes with Repeat HgbA1c E c usn Bad DiabetesWRepeatBadF2 E Methods c USR CHF Primary March 2012 Discharge E c USR Inpatient Admit w CHF in Mar 2012 E c usn Pneumonia Females 2010 Admit Primary E c USR Pneumonia Females 2010 Discharge Primary E c USR Pneumonia Males 2010 Admit Pri
85. ow F FF F 9 e EHI GEIL IIR UI Left click on the drop down so we can choose the Analysis Definition 67 Study Designer A Run All _ Save As Im an 9 x Start Us Blackout 0 Days z d H T B Method Name Adv TTO Admit Method Description Time To Outcome Estimation Points 30 60 90 180 365 Graph Name Ady TTO Admit Resulting in We Analysis definition Y EAS K Adv TTO Admit X Run Method Show Results USR Admit O0 365dAfter USR Admit 0 365dAfterDeadWhenIn USR Admit 0 365dAfterDeath USR Inpatient Admit USR InpatientWAtributeDeath Days 68 ETDE Jc Run all saveas a Gex x 4 gt 4 wf TTO admit f Adv TTO admit X Start Blackout 0 Days B Method Name AdvTTOadmit Method Description Time To Outcome Estimation Points 30 60 9080 365 1 1 1 Days Graph Name KM Graph Name Now rename the graph Advanced TTO First Admit ETE amp Run all Osven 5x x 4 gt wf TIO admit vf Adv TTO admit X Run Method un titans Analysis definition 5 Start USR AdmissionFirstO 365days Blackout 0 Days un D o r o adii E F Time Units Days B To Simple Method Name Adv TTO admit Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name Advanced TTO
86. p E O 1 1 10 Test Clinic 0 1 1 4 EventDateTime 0 1 1 11 Test MD 0 1 1 13 Lab Test Type Q 0 1 1 12 True First 0 1 1 1 Count i 0 1 1 13 Lab Test Type E f 0 1 1 9 Lab Test Value E Restore Default Notice that Lab Test Value in Available Attributes is grayed out Lab Test Value is visible in Selected Attributes 138 Run the Method i i i Save E 0d Ever Ton U DOE UAEeWHEanD FimECpoherctDuraEan W P DOLI Feat WOE a OL Coast E 01 1 2 Exe t D D 0 1 3 3 pEwentT yee D O14 c Eventa jp 0 1 1 5 i Ere nitama 4 OLI oCetegeric Lab Tasi Value S 0 1 1 7 Grnuped Normaeric Lube OLLE Lab Taxt Gr 01 1 9 lab Test Va n 1 1 10 Tant Clinic 42 0 1 3 11 i Test MD OO True First io 0 11 13 ple Tasi E i dE Run AM C savea lee XC ETITIPEETIDISTIT 3 t i Patsentil 3 n 2 unn D 0 3 ses naa ethnicey 3 15 cande D 0 40 i Prelerresd Language 10 20 Race 1 0 23 Age at Index Date iy 63 IEvectDabaTima jf G0 40 IDinpcaitian jf D 41 Inpatiant IEDI3 Group i 6 0 13 i APR DR Eb O 1 1 4 vmciDataTi ma y 0 1 1 13 Lad Test Type TOLLI Gout E 1 1 0 Lab Task Value 139 A Study Designer a Run All C Save as a 9X X Coo oOo o z m wf TTO Admit W Adv TTO Admit TTO Adv When In V BasicWbcEcholist X 3 wg 5 1 Reset Method Run Method Show Criteria fy S List p o 5 Prvo
87. p 3 sES FJ D 14 Ethnicity D D 15 Gender r p 19 Preferred Language 55 p 20 i Race 3 D 22 Age at Index Date G 0 3 EventDateTime amp 6 0 40 Disposition 5 6 0 41 Inpatient ICDS Group Run the method 121 Study Designer A Run All C save as X Main Groups Methods T l wf TTO Admit M Adv TTO Admit V TTO Adv When In V BasicWbcEcholist X Reset Method Run Method Show Criteria List Pivot Show Identifiers mo 4 Page p of 106 b Im Total records 1683 10 Male Disch 1Disposition SUB ACUTE 27 L CARE amp HOM 25 fections amp Inflammations Notice APR DRG and Disposition To return to Study Designer criteria page hover over Show Criteria and left click Study Designer p3 Run All C save as p pl pM X Main Groups Methods 4 V TTO Admit v Adv TTO Admit v TTO Adv When In Reset Method Run Method List Pivot C Show Identifiers M44 Page 1 of 106 gt p Total records 1683 APR DRG qual GROUP 1 INDEX Pneumonia 2010 Male Other Pneumonia REGULAR HOME SELF CARE 01 Other Pneumonia REGULAR HOME SELF CARE 01 Other Pneumonia SNF INCLUDES LOEB SPECIFY 03 Export GROUP 1 Il 2 26 2010 10 3 2010 3 23 2010 10 26 2011 3 2 2010 5 1 27 2010 1 30 2010 6 22 2010 12 8 2010 4 29 2010 9 8 2010 2 2 22 2010 10 25 201t 2 13 2010 4 10 2010 2 15 2010 122 d Run All Save As 9x
88. pe z EVENT amp SelMenuOptionz ModifyEventDefinition amp EntityID ewEventDefinition amp gui Name EV C332c Please Select v End Point Event Please Select X Update Update and Close Event Definition Options Name the definition Cardiac Echo Choose Echo Date from the Event Type drop down menu C G Study Designer Assessment Functional IADL is Cancer Registry First Recurrence Date Cancer Registry Initial Diagnosis Date n Cancer Registry Site Spec Factors Init Dx Analysis Canvas CPT4 Procedure Date Date of Birth Management OX CRB Manage Analysis Er Death In House Death Date Death Social Security Death Date Death Social Security or In House Death Date Echo Date a E Event Definition Palette Webpag Cygischarge ED Seen e https clg emerginghealthit com CLGNET Modules Ever ED Triage pe EVENT amp SelMenuOptionz ModifyEventDefinition amp Entit Findings i J EE M Immunization Date Event Inpatient Admit T Inpatient Discharge End Point Event Please Select ype Internal MedAdminDate update Een Lab Test N Med Order Start Med Order Stop Medication Allergies Start Date Medication Allergies Stop Date Microbiology CollectDate Non Med Allergies Start Date Non Med Allergies Stop Date z Name Cardiac Echo IL r Event Definition Options Update and Close 144 ELAD f CardiscEcheEF180DaysBefore
89. pens Click on the methods tab to create a new method CiG Study Designer 1 studies PO Study Designer b Run AII Save As s s USR Admission Repeat Diabetic E 7m sE USR Diabetes contr al outcome Study Name Pneumonia Readmit amp s USR Pneumonia a Description Enter Description i E USR Pneumonia Readmit B ul E z dne Group Name Cohort Collection a Pm 1 J Male Pneumonia c usn Pneumonia Males 2010 Discharge Primary E 3 2 Female Pneumonia c usr Pneumonia Females 2010 Discharge Primary E Add a new method by clicking on the plus sign ETT GS Runan O save as a lex x E TTO admit X mam Show Criteria 2 3 Demographics rarget Event All Events Patient List 5 p s Male Female Unknown le Disi 0 1 Male Discharge 3 baseline 753 100 0 0 0 0 Female Discharge 0 0 927 100 0 0 77 The new method dialog box will open New Method Dialog Method Type Select a Method Type M SGU Method Name Method 2 Add Left click on Select your Method Type and choose Time to Outcome New Method Dialog Method Type Select a Method Type v VEUT Method Name Select a Method Type Time to outcome List Time in Range Add The resulting image is New Method Dialog x UEM Method Type Time to outcome x New LELE Method Name Method2 eee Add j Canc
90. rral Agency DRG Drg Weight Admit Attending Admit Housestaff Discharge Attending Discharge house staff Facility Insurance Plan IPA Insured Line of Business M Toadmit e Adv TTO admit BasiclistForCohort X Selected Attributes G 0 1 GroupNo 133 D 1 PatientID f D 2 MRN G 0 38 Disposition G 0 12 DRG Restore Default Methods Groups Main VvvVVY VY AAAAASA T ES IK TTO admit amp Run All SaveAs pox X j Adv TTO admit Reset Method Run Method Show Criteria List Pivot _ Show Identifier 1 M4 4 Page 1 GroupNo M mM Rom i RA RR RA RR RA RR PatientID 13177 15511 23278 23839 24380 24794 32541 40330 40826 48595 50620 51210 me mm M BasicListForCohort X of 103 gt ppl Total records 1646 MRN Disposition clg 13177 REGULAR HOME SELF CARE 01 clg 15511 REGULAR HOME SELF CARE 01 clg 23278 SNF INCLUDES LOEB SPECIFY 03 clg 23839 HOME CARE SPECIFY 06 clg 24380 SNF INCLUDES LOEB SPECIFY 03 clg 24794 SKILLED NURSING FAC SHORT TERM SUB ACUTE 27 clg 32541 SNF INCLUDES LOEB SPECIFY 03 clg 40330 REGULAR HOME SELF CARE 01 clg 40826 REGULAR HOME SELF CARE 01 clg 48595 HOME CARE NON HOSPICE PERSONAL CARE amp HOM 29 clg 50620 SNF INCLUDES LOEB SPECIFY 03 clg 51210 REGULAR HOME SELF CARE 01 Lj me am memes nm fo rmt ttm memes nani na DRG SIMPLE PNEUMONIA amp PLE
91. rt When you left click on the downward arrow a drop down menu appears Choose excel 2007 as excel 2003 is limited to 65 000 rows and a smaller number of columns Export Method Export Type Excel2007 v _ Export Method Export Type Excel200 Y You can download your exported files here Data file 30 Left click export and a data file appears Left click on the data file and the spreadsheet opens Ed S c 793de582 1264 4ad8 8255 17ab317aa748 Read Only Excel FILE HOME INSERT PAGELAYOUT FORMULAS DATA REVIEW VIEW ACROBAT E e m FLLIJ E d cut Arial 10 AA B v E Wrap Text General ia I m EB Copy l nant p HH Paste S E tuse us z lt Z Ss EE 2 0 00 Conditional Formatas Cell Insert L 3 W Format Painter BIU E 2 A zzi esx uisi arcs 96 9 an Formatting Table Styles Clipboard rs Font fa Alignment p Number ra Styles A1 M f A B C D E F G H J K E M Summary Name Pneumonia 2010 Male discharge dc alive w UT Created By ebellin7 vem m Modified On 2015 03 26T22 33 00 D e 5 03 26122 33 Distinct MRNs 741 P Modified By ebellin7 Type Cohort Built On 2015 03 26T22 33 00 INDEX EVENT Definitions list Created On 2015 03 26122 33 00 Earliest Of Any Pneumonia InpatientDischargeDate All ENIM discharge All Of Pneumonia InpatientDischargeDate WHEN IN 2010 WITH Male
92. rward for a defined period and calculate the percent of time spent in prespecified ranges of laboratory test values For example you could follow the INR values of a cohort of patients from their index date for a year and summarize the quality of their anticoagulation by the time spent in the ranges of 1 1 9 bad low 2 2 8 good 2 9 infinite bad high In this Basic CLG literacy course we will cover the first two analytic patterns leaving to you in self study video tutorials and manuals to learn the third analytic method During this course watch 1 7 For study designer 8 9 E E E d di tu Object management Time to Outcome Introduction Time to Outcome Advanced List Introduction List Intro 1 List Intro 2 List Advanced The following are optional for advanced students TIR lecture see if lecture is adequate and if it needs to be enhanced TIR Introduction 10 TIR Advanced argeDate WHEN IN march2012 x bialSecurity SSAorInHouse WHEN IN c Time to Outcome 1 Tutorial 14 Study Designer Time to Outcome Simple Method https www youtube com watch vZMw3XVDeWFHw amp index 14 amp listZPLf7raPnmlLOeAWU2cNf 2jIMDRqQuUTCum Let s consider a readmission question For all the patients with pneumonia from the moment of discharge alive how long did it take for the patients to get readmitted Readmission for any cause The method required is a time to outcome analysis First let s open up study d
93. s a GUI editor below 96 Bl Edit Selected Admit 0 365dAfterDeadWhenln r Analysis Canvas E AD S Admit 0 365dAfter Earliest of Admit_0_365dAfter And ay Admit 0 365dAfter All of Inpatient Admit InpatientAdmissionDate within 0 to 365 Days After Cohort Collection Date l AND Alte Eventi New Event Definition when in New Duration Definition with New Demographics Definition r Edit Selected cendition Eventi E Ng ga Update 1 Put in a new name Death in House or in Social security 2 Choose All of A bug in the program requires death to always be addressed with All of and not earliest 3 Left click on downward arrow New Event Definition to create a new event 97 Edit Selected Condition Eventi NOT All of WHENIN Cohort coi NewEventgMf fographics v k Select from Update Select from Common Defini ion rom CLGHET Modules EventDefiniionBulider EventDetsinmonBuilder ssp Def Types PENTR HMen phon NesEventOefintion amp Enta Name EV C6503 A Please Select End Point Event Please Select We want to find the event Death We will use Death in House or Social Security 1 Rename your event type to be DeathinHouseorSS 2 Find the event type from the drop down menu Cancer Registry First Recurrence Date Cancer Registry Initial Diagnosis Date Cancer Registry Site Spec Factors Init Dx CPT4 Procedure Da
94. sionFirstO 365days E USR Admit O0 365dAfter E USR Admit O0 365dAfterDeadWhenIn Bi E E USR Admit O0 365dAfterDeath E USR Inpatient Admit Method Mame TTO Adv When In FE USR InpatientWAtributeDeath Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name KM Graph Name Now select Usr Admit O 365dAfterDeadWhenln i Study Designer Main Groups Methods d Run All Save As X x 4 E37 TTO Admit Lv Adv TTO Admit TTO Adv When In X Run Methad Outcome Analysis definition USR FirstAd mission S65days F USR Inpatient Admit FE USR InpatientWAtributeDeath Method Mame TTO Adv When In Method Description Time To Outcome Estimation Paints 320 650 90 180 365 Days Graph Name KM Graph Name 103 Name the Graph DeathAfterReadmissionTTO Py Run all saveas hl 9X X 4 Y Ime TTO Admit Wf Ady TTO Admit f TIO Adv WhenIn X Study Designer Run Method Show Results Outcome Analysis definition USR Admit 0 365dAfterDeadWhen un H T Be To Simple Method Name TTO Adv When In Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name DeathAfterReadmissionTTO Run the Method Study Designer A Run All Saveas h a X X Ra Adv TTO Admit TTO Adv When In X Outcome Analysis definition rd USR Admit 0 365dAfterDeadWhen Admit
95. sk Relative Risk Risk Difference Male Pneumonia 793 243062 etd quu eee UM Female Pneumonia 1 338 baseline 1122 306170 265 1 411 UE PE EA X TTO Admit tun Method Demographics Target Event All Events Patient List Number Target Events Achieved Person Days at Risk Incidence Density per year of risk Relative Risk Risk Difference Male Pneumonia 793 243062 apne Roast Pee aera Female Pneumonia 1122 306170 je uu baseline 1 411 Note that can be skipped on first reading The All events tab in study designer reruns the Analysis definition but changes the index line of the Analysis definition from Earliest to All of Since the first condition line on the event canvas is All of the Advanced and Simple mode yield the same result in the All events tab Should the first condition line be Earliest then the Advanced Mode would only count the first admission for each person while calculating all the follow up time for the incidence density calculation Let us re review what an analysis definition is in the context of Time to Outcome Method The analysis definition attaches an event to the cohort The analysis definition counts time from the cohort s index date until the event of the analysis definition 72 In our example the analysis definition attaches the first admission after cohort s discharge date to the cohort In our example the analysis definition measures elapsed time from the cohort s discharge date to t
96. st proximal to the start date will be the latest event in this window The power of the Advanced Method in terms of time is that it is able to select the latest or the earliest event in any time window The Simple method can only detect the most proximal which means from the perspective of the calendar Earliest in the forward direction Latest in the backward direction In the Advanced Mode the outcome event itself can be complex with multiple events or multiple temporal constraints as we shall later see 56 We will now build the same study in Advanced Mode that we built in the Simple Mode using our pneumonia groups Instead of modifying the current TTO admit method we will add a new method and a new analysis to this study by Left clicking on the plus sign You are in Study Designer Click the plus sign to add a new method a ow a e ue v a I rw Q SMIDE Py Run AM O Save As p 9x 3 _ _ Left click to open new ali Method Run Met a Show Results z iris Analysis definition Start ass Blackout 0 Days F Time Units Days B To Simple Method Name TTO admit Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name Readmission for M F Pneumonia Pts 57 This yields Method Type Select a Method Type z GUTE Method Name Method 2 Cancel Choose as
97. stro outpt visit CAD tme to gastro J serf User arai AD usn Admit 0 365dAfter j USR Admit_0_365dAfterDeath We are now going to take advantage of work we ve already done by modifying a previously built AD and saving it with a new name so we end up with both the original AD and a newly named AD Hover over the Admit O 365dAfter and right click and edit This is the original AD you built a while back that did not have death as an attribute of the admission Analysis Canvas 94 Analysis Canvas 4 Management Ab Manage Analysis Sw Bl system Analysis LAD Bj time to gastro iser LAD USR Admit 0 36 AD B USR Admit 0 36 USR Inpatient A The following image appears Analysis Canvas EIERS T SaveAs 229 X X Name amp dmit Q0 365dAfter AD B time to gastro Ap B Admit 0 365dAfter Earliest of Any And L5 Admit_O 365dAfter All of Inpatient Admit InpatientAdmissionDate within 0 to 365 Days After Cohort Now we will concentrate on the analysis canvas 95 Analy WSS Canvas SCAD Bada sacintent paiet USC ANN DELI 3 1 Check off save as so you can create a newly named AD without overwriting the original 2 Rename the AD as Admit O 365dAfterDeadWhenln 3 Right click on the index line to show drop down menu and choose When In EIER uva Name Admit 0 365dAfterDeadVWhenln Sub Group This creates a new When In condition line and open
98. t 7 Show Identifiers Export m 4 Page 1 of 106 P IT Total records 1583 InFirstCohortDuration First WBC 1Lab Test Type Methods TY 24444 a 4 LI Study Status The spreadsheet we generated first included APR DRG and disposition from the cohort s index event We have just added the White Blood Counts and Lab Test Values during the hospitalization Now we want to add the latest Cardiac Echo ejection fraction in the 180 days before the patient s discharge cohort index event Let s build the analysis definition to find the Cardiac Echo ejection fraction To get the echo ejection fraction you must get the echo event which is known as EchoDate To find relevant event types like cardiac echo go to the Event Manual 3 v2 6 stored within your Box link Use the find Ctrl F capability of Adobe Reader to search for the words cardiac echo 140 Or alternatively do something novel and read the table of contents in the Event Manual 3 v2 6 where you can find event types Let s return to Study Designer Show Criteria page Hover over Show Criteria and left click j Study Designer d Run AII C Save as V X X 1 F v TTO Admit v Adv TTO Admit v TTO Adv When In y BasicWbcEchoList X Reset Method Run Method un S List Pivot wo a r Show Identifiers Export E eC 4 Page 1 of 106 Ld m Total records 1683 E k Ji d l Ji J t imma oo Bl c Mabon tT RAMAI CETE Et bhm sm
99. t Na analis definition selected a a Bisckoor d Dares The Analysis Canvas opens to display the Management pane on the left and the Analysis Canvas on the right CiG Study Designer Analysis Canvas i Management O eee save as ul X X al Sys Bj system analysis i AD B sartea gastro cutpt vin A Analysis Canvas AD By time to gastro CAD Dh New Analysis 7da025b All of Any And a Psar user Analysis An B use Adm 0_365dAfter USR Admit D 365dARerDeadVWhenIn P An Bj USR Admi 0_365dAfterDeath d USR inpatient Ades Enter the name WbcWhenlnFirstCohortDuration Since we are interested in the White Blood Count s during the original hospital admission we need to add a When In duration condition Right click on the index event line to open the menu Left click When In 125 EESE saeas 521 29 X X Within 12 Sub Group 126 Resulting in EM AD JN New Analysis 7da025b All of Any And MA Eventi New Event Definition when in New Duration Definition with New Demographics Definition F Edit Selected Condition Event1 MOT Al of New Event Def WHEN IN New Durston De v WITH No Demooni v Name the Event1 First WBC Since we want the first white count during the hospitalization change All of to Earliest of We want to find Wbc Count Left Click on the downward arrow of the New Event
100. t Wbc condition line to open the menu Right click on Edit 160 Name WbcWhenInAllCaohortDuratioan EESE F saveas 21279 X Analysis Canvas B nc A Ene In the Gui editor below change the name to All WBC Change Earliest of to All of AES F SaveAs 229 X X Name WbcWhenlnAllCahortDuration EI AD JJ wbewheninFirstCohortDuration All of Any And EE First WEC Earliest of WBC LabTestDate WHEN IN Cohort Collection Duration Edit Selected Condition B Edit Selected Condition All WC PINOT WBC LabTestDate Y WHEN IN Cohort Calection Duration Y WITH No Demographics Update Update and Close 161 Update and Close and the AD looks like this AER Osven OER Name WbcAllWhenInCohortDuration Analysis Canvas E AD B wocatwhenincohortDuration All of Any And L Ei All WBC All of WBC LabTestDate WHEN IN Cohort Collection Duration Notice the All on both the Index Event Line and on the first condition line Save and close The new AD appears in the management pane Analysis Canvas Management ox rapi Manage Analysis 4 Svs B system analysis LAD B earliest gastro outpt visit AB B time to gastro User user AD B usr Admit 0 365dAfter i AD 5 USR Admit 0 365dAfterDesdWhenin CAD B USR Adm 0 365dAfterDesth e Inpatient Admit CR Bun
101. tcome 5 Start End un Blackout 0 Days i Time from Start 365 8 F Admission e D Time Units Days B To Advanced Method Name TTO admit Method Description Time To Outcome 2 Estimation Points 30 60 90 180 365 Days 3 Graph Name KM Graph Name 4 All the following umbers refer to their counterpart in the above graph 1 2 accept the defaults 3 Estimation days seem appropriate 30 60 90 180 365 days This will build a table of cumulative incidence first admission by day 30 60 90 4 Rename the graph name to something that will be meaningful 20 EXE runan 05d 9x x W gt TTO admit X rE Run Method 2 un a o Outcome T Start End un Blackout 0 Days Time from Start 365 Admission o P por E D Time Units Days B To Advanced Method Name TIO admit Method Description Time To Outcome Estimation Points 30 60 90 180 365 Days Graph Name Readmission for M F Pneumonia Pts 4 1 The Graph Name is renamed 2 Left click on Run Method 3 Left click on Save Note In event canvas a build automatically saved the rules and the cohort or event collection warned you not to manually save after an event canvas build as a user initiated save in event canvas saves only the rules and deletes the mrns The convention in study designer is the opposite A run in study designer is not automatically saved A save or Save and exit
102. te Date of Birth Death In House Death Date Death Social Secunty Death Date gt Echo Date 98 Update and Close Edit Selected Condition Eventi FINOT All of DeathInHouseOrss SSAorInHouse WHEN IN New Gur Def v WITH Vo Dares v Now we are going to choose a New Duration Definition Let s look at our options New Duration Def Select from i Copy from d SESE D Select System Duration gt 1NOT Ez Cohort Collection Duration gt SAorInHouse WHEN IN Cohort Collection Duration Y WITH No Demographics Y mmc New duration would create a new calendaric interval Select from would select from the calendaric interal Copy from would copy the calendaric interval Select System Duration We want our death event to occur during the duration of the hospitalization in condition line one We therefore choose Select System Duration and accept Looking Glasses offer of First readmit Inpatient Stay 99 Cohort Collection Duration would point outside the analysis canvas to the event used to build the cohort cohort member s index event Since we want the death to occur in the downstream admission we avoid this option For the duration you choose Select System Duration from the analysis canvas CLG offers the only condition line on the canvas with a durational event that you can select Note Select System Dura
103. te USR monoclonal gammopathy with bm SG B abader No MI patients CABG V4 fc536 S s USR Pneumonia Readmit USR readmissionHope Left click on the S of the study Usr Pneumonia Readmit have built Usr is the prefix given to the studies that you 51 CiG Study Designer s s USR CABG in no MI group s abader CABG in no MI group 7e9d5 S s USR CABG in no MI group redone sE USR cancer si USR Cancer Incident Resource consu sl USR cancer Predict BE USR cancer Predictbad BE USR Compare follow up two years s s USR Diabetes control outcome sl USR Diabetic siS USR Discharges outpt follow up sE USR MedicaidPopAdmitRate s s USR MedicaidReadmitNoOutptRequired sE USR MedicaidReadmitRate s s USR monoclonal gammopathy with bm s abader No MI patients CABG V4 fc536 USR Pneumonia Readmit P XS X 1 j 2 n 3 iissionHope 4 Taek TraininaCHF Four images appear 1 A pencil meaning edit allowing you to open the study run it or modify it 2 Share symbol lets you share your study with a colleague mnemonic open arms sharing 3 PDF this will provide a pdf summary of your study without identifiers 4 Red X if you left click on this the study will be deleted and unrecoverable Left click on the pencil and the study designer canvas opens to the right of the Management Pane 52 Study Designer Studies iad ee Xe kan Sees pl Lx MC g steer cana in rof group eats a a mca
104. temporal operator pointing from the Analysis definition to the Cohort s index Event e The last ejection fraction from the latest Cardiac Echo performed prior to the hospitalization in the preceding 180 days Best LV ejection fraction Because you now want to obtain attributes from events not belonging to the index event White Blood Count and Echo Ejection Fraction you must now select what are called Outcome Event Attributes If there were only one outcome event and it was the Earliest outcome event then you could use the racetrack paradigm in the simple mode Because you want more than one outcome event in our case WBC s and Cardiac Echoes you must go to the Advanced Mode and build two Analysis Definitions one to be used to collect the Earliest WBC during the hospitalization and the other to detect the Cardiac echo within 180 days of admission 115 Even if you had wanted only the White blood count event you would have needed the advanced method to use the When In duration Temporal operator The Simple method only permits an implicit Within time operator We will now proceed to use the Advanced Method so we can look at all these three events at the same time in one spreadsheet Open a new method in Study Designer Reopen the Pneumonia Readmit Study CiG Study Designer ETTEN C 5 USR Admission Repeat Diabetic 5 ls USR Diabetes control outcome s USR Pneumonia USR Pneumonia Readmit
105. the question that you plan to answer See Riddles in Accountable Health Care Chapter 5 Heads or Tails which End is up 49 Time to Outcome Advanced Method First Recapitulate Results of Simple Method with Advanced Method Tutorial 17 Time to Outcome Advanced Method https www youtube com watch vzd8i9XocczLO amp indexz17 amp listzPLf7raPhnmILOeAWU2CcNf2iIM DRqQuUTCum To transition from the simple method to the advanced method in study designer we will first redo the pneumonia study you performed in the simple method using the advanced method Once we have used the Advanced Method we will compare the results with the Simple Method and show you where they are the same and where they are different We will then proceed to evaluate a problem for which the Advanced Method is necessary to solve the problem You are familiar with the racetrack paradigm of study designer Let s open up our old Pneumonia Time to Outcome Study Go to Study Designer The management pane appears 50 pm USR CABG in no MI group S B abader CABG in no MI group 7e9d5 USR CABG in no MI group redone G Is USR cancer USR Cancer Incident Resource consu S Is USR cancer Predict USR cancer Predictbad I s USR Compare follow up two years USR Diabetes control outcome s g USR Diabetic USR Discharges outpt follow up sli USR MedicaidPopAdmitRate USR MedicaidReadmitNoOutptRequired TEE S Is USR MedicaidReadmitRa
106. thod Name Time to e us List Time in Range Add Cancel 1 Name the new Method TTO Adv When In Click Add New Method Dialog Method Type Time to outcome eile Method Name py Add Cancel A new method tab appears on the top of the canvas adjacent to the previous methods that you ve already built 90 Study Designer A Run All O Save As MEN x 4 5 ER sro Admit Ra Adv TTO Admit TTO Adv When In X essere ee Run Method Outcome ui No outcome E selected L 7 a Ta Advanced Method Name TTO Adv When In Method Description Time Ta Outcome Estimation Points 20 60 90 180 365 Days Graph Name KM Graph Name We need to move to advanced mode from the simple mode Hover over To Advanced and left click By Rum all C savesas W V5 X x T ES V TTO Admit Adv TTO Admit TTO Adv When In X d Study Designer Run Method Outcome Mo outcome selected The Racetrack Metaphor changes from Simple to Advanced mode 91 p3 Run All Saveas p X X Study Designer E 4o M TTO Admit A Ade TTO Admit TTO Adv When In X Run Method SSS Analysis definition m d End No analysis definition selected tt un 1 i a 1 2 F Time Units Days To Simple Method Name TTO Adv When In Method Description Time To Outcome Estimation Points 30 60 90
107. thod Name BasicWbcEchoList Ly Method Description List Selected Attributes D v 3 ses 3 p 14 Ethnicity D D 15 Gender D p 19 Preferred Language D D 20 Race 3 p 22 Age at Index Date ty G 0 3 EventDateTime Outcomes Demographics Group jt G 0 40 Disposition amp G 0 41 Inpatient ICD9 Group amp G 0 13 APR DRG E Restore Default a Study Status a Note there are no Available Attributes in the Outcomes tab We want to add a column of white blood cell counts to our list using the Analysis Definition we ve just built We need to add the Analysis Definition WbcWhenlnFirstCohortDuration to our Advanced mode racetrack When we do that all the attributes of the added Analysis Definition will appear in the Outcomes tab of the Available Attributes and we will be able to select White Blood Count Hover over the dropdown menu and choose WbcWhenlnFirstCohortDuration Note when you first build an Analysis Definition you will find it at the bottom of the dropdown list In the future you will find it alphabetically arranged 135 Study Designer d Run All O Save As uw X x 4 y Ie TTO Admit v Adv TTO Admit iv TTO Adv When In K BasicWbcEchoList X Reset M hod Run Method Show Results eT ts Analysis definition No analysis definition selected vu USR Admit 0 365dAfter USR Admit 0 365dAfterDeadWhenIn USR Admit O0 365dAfterDeath B To c USR CardiacEchoEF180Days
108. tion is different from the last choice in the drop down cohort collection duration which reaches off the analysis canvas and points to the cohort member s index event Imagine if you had chosen to point to the cohort member s original event you would be requiring that the patient died during the first pneumonia hospitalization which is impossible because you are assuming the existence of a subsequent admission And we do not readmit dead people The Gui editor now looks like A Edit Selected Condition Death In House ar Sacial Sec ENOT All of Death in house or 55 5S5AorInHouse WHEN IN Admit 0 365dAfter Inpatient Stay Y WITH ENTE Update and close Update and close M Save As Name Admit_0_ 365dAfterDeadWhenln P l Analysis Canvas E AD Admit_0_365dAfter Earliest of Admit_0_365dAfter And vag Admit 0 365dAfter All of Inpatient Admit InpatientAdmissionDate within 0 to 365 Days After Cohort Collaction Date iw AND LR Death In House or Social Sec All of Death in house or 5 SSAorInHouse WHEN IN Admit_0_365dAfter Inpatient Stay 100 Save and close Analysis Canvas Save As Us Name Admit_0_365dAfterDeadhvt C Analysis Canvas Lap B Admit 0 365dAfter Earliest of Any And i peer Admit 0 365dAfter All of Inpatient Admit InpatientAdmissionDate within 0 to 365 Days After Cohort amer cns i AND USR Admit 0 3e5dAfter Eu Death In
109. tion Options 128 e https clg emerginghealthit com C LGMET Madules EventDefinitionBuilder EventDefinitionBuilder aspx Def Typezl Name WBC Event Type Lab Test Update Update and Close Event Definition Options We want Lab Test to be a White Blood cell Count Hover over Definition A box appears around the word Definition Right Click on Definition Left Click on add Condition Event Type Lab Test Update Update and Close 129 Dpage log Definition Canvas Y A https clg emerginghealthit com CL GMET Modules EventDefinitionBuilder EventDefinitionBuilder aspx DefTypez EVEN NOT Please Select Please Select Update Update and Close Choose from dropdown menu Lab Test Type rr Ez and Class NOT Please Select Please Select Ea Select z Close egoric Lab Test Value Numeric Labs Lab Test Type NOT Earliest WHEN IN Coh 130 Choose 2 Definition Canvas Webpage Dialog NOT Lab Test Type Choose WBC VOLUNTARY TEST IN NY VOMITUS OCC BLOOD ysis Canvas ee VWF Ag Multimer Ana VWF ANTIGEN VWF COLLAGEN BIND VWF Interpretation ml VWF MULTIMER STUDY E Definition Canvas Webpage Dialog VZ Ab CF VZ Ab IgG EIA e https clg emerginghealthit com CL GNET Modules Event VZ AB IGG IFA p VZ AB IGM EIA EE M VZ AB IGM IFA VZ CSF CF Ab Name WBC Event Typ
110. to use the triple button in study designer to open event canvas and build new cohorts We are going to repeat the previous study but switch the index Event line from the second condition line to the first condition line CiG Study Designer a EMENEZEE ex osx s j USR Diabetics with p control hosp s g USR Pneumonia Readmit Grp ID Baseline Group Name Cohort Collection 1 e Group Name 1 No cohort collection selected la x z 5 X in If you had not already built the necessary cohort you could use the triple button to open event canvas and built a new cohort 44 CiG Study Designer j Management SE EN save as E E m Fae Name BadDaveesiRepeateoori 4 EE Collections and Cohorts EH Pneumonia Bir Bus Pneumonia Males 2010 Dit hagge Primary LC Buss Pneumonia Females 2010 Dischard armar B usn Pneumonia Females 2010 Admit Primary e c usn Pneumonia Males 2010 Admit Primary bd E CHF J USR CHF Primary March 2012 Discharge be usn Inpatient Admit w CHF in Mar 2012 Ewent Canvas run eee Having built the cohort click on the button to return to study designer Lc USR Bad Disbetes Once having built the cohort you could then left click on the management pane of the event canvas and find yourself back on the Study designer canvas Fortunately we have already built cohorts with the focus on line 1 They were saved as Ba
111. udy Designer ox Run All _ Save As Grp ID Baseline Group Name Cohort Collection 1 Male Pneumonia 1 USR Pneumonia 2010 Male discharge x Main ps 1 Type in Male Pneumonia 2 Left click on to add another group Study Designer Jy Run All O savas Xj 9X X m Main Grp ID Baseline Group Name Cohort Collection e Male Pneumonia USR Pneumonia 2010 Male discharge e X C Female Pneumonia c USR Pneumonia 2010 Female dischal x This ss chooses male as baseline The way we have set this up we have made males as the baseline But want to compare the relative risk of admission of males to females so have to change this to Female baseline by clicking on the baseline button near the female cohort 11 Study Designer P Run All L SaveAs w BM X m c Grp ID Baseline Group Name Cohort Collection AO Male Pneumonia i i o 1 amp USR Pneumonia 2010 Male discharge X 1 o 2 Female Pneumonia USR Pneumonia 2010 Female discha X i Females are now baseline 2 Left Click on Methods Methods We have successfully chosen the two groups for study the male and female pneumonia cohorts of 2010 We have successfully established that the females will be the baseline We have successfully named the groups so any table or graph created by study designer will be clear We have finished with the Groups canvas of Study
112. up a O D 12 Patient Address p 13 Date of Birth omes Demographics Group 3 D 14 Ethnicity Restore Default Study Status 4 Note the large number of greyed out Available Attributes We want to remove unnecessary selected attributes so that the only remaining demographic attributes are SES Gender Race Ethnicity Age at Index and Primary Language Hit the Crtl left click to highlight unwanted demographic Selected Attributes Demographic attributes are preceded by a BIG GREEN D DO NOT remove the Selected Attributes from the group in BLUE amp l Those attributes used as conditions on the cohort s index event are automatically brought over into the selected attributes in BLUE 119 Now we will remove the unnecessary demographics Left click the red x View Wide Long m 1 3 D 14 Ethnicity 9 D 15 Gender D n 19 Preferred Language m D 20 Race D rm ha TI FUR CR D D 22 Age at Index Date SPE er a Serr Tee See see eee EE LETT SEE See ee Tee seer Se ee eT ae f m myr Bi 1 View the list of Available and Selected Attributes view 9 wide Available Attributes Selected Attributes E i D D 1 PatientID G 0 1 GroupMa D 2 MRN D 1 PatientID Demographics Outcomes Demographics Grou D 3 SES D 2 MRN vidi aid Removed removed 3 D 4 Apartment demographics D D 3 SES are no
113. utes lt gt Bu M Adv TTO Admit _Reset Method Run Method Jy Run AM Save As e 9 X X KA TTO Adv When In BasicWbcEchoList X 3 3 5 CEEE 0 0 1 EventName 0 0 2 EventDateTime Methods Group 0 0 3 Eventio 0 0 4 EventType 4 M Q WbcWheniaFirstCohortDuration X 0 1 1 2 EventID 9 0 1 1 3 EventType Outcomes Demographics 5 0 1 1 4 0 1 1 5 Eventtiame 0 1 1 6 Categoric Lab Test Value o 0 1 1 7 Grouped Numeric Labs 0 1 1 8 Lab Text 0 1 1 9 Lab Test Value Q 0 1 1 10 Test Clinic Restore Default This results in 4 R 0 1 1 First WBC View Wide Long lel C G 0 1 ou 0 2 3o 3 ses 3 0 14 ethnicity ofa oun Language PTJ 0 20 Race 3 0 22 Age at index Date G 0 3 EventDateTime G 0 40 Disposition G 0 41 Inpatient ICD9 Group G 0 13 APR ORG 0 1 1 4 rEventDateTime O 1 1 13 Lab Test Type View Wide Long W Available Attributes Selected Attributes es a 0 1 1 3 EventType f p 22 Age at Index Date e 0 1 1 4 EventDateTin G 1 1 5 EventDateTime a 0 1 1 5 EventName A G 1 1 39 Disposition Z 0 1 1 6 Categoric Lab Test Value amp y G 1 1 40 Inpatient ICDS Group amp 0 1 1 7 Grouped Numeric Labs amp G 2 1 5 EventDateTime 7 8 0 1 1 8 Lab Text amp G 2 1 39 Disposition E 0 1 1 9 Lab Test Value amp G 2 1 40 Inpatient ICD9 Grou
114. vice Limited data sets under Hipaa must be treated with respect and encrypted if not onsite in a monte machine 1 clicking on show identifiers will result in a Hipaa challenge demanding that you justify your request to view identifiers 2 exports data in either excel or csv format The Hipaa challenge looks like this 28 HIPAA Validation Identifiers You have privileges to see patient identifiers on this patient listing To see identifiers click the checkbox below and indicate your authorization accordingly To see the report without identifiers click on the OK button below IV Show Identifiers Please Indicate your QA IRB Approved 5 Patient Q CLG authorization Project Research Worklist Validation All requests for data with identifiers are subject to audit QA project you have registered your project with the QI department IRB research enter the IRB number that permits you to view identifiers Patient Worklist you are actively using these lists now to deliver clinical care to patients for whom you have a physician patient relationship Example Creating a list of patients who have not received the influenza vaccine so you can call them in for vaccination CLG validation rarely invoked you are working with the CLG maintenance team to help determine the nature of a bug in the program When you left click export the following appears 29 Export Method Export Type Excel2003 v Expo
115. w here 3 5 city D p 14 Ethnicity p e state J p 15 Gender 3 D 7 CensusTract 33 p 19 Preferred Language 3 n 11 p v 12 3 v 13 M D114 153 D 15 p 16 Latitude Longitude CensusBlockGroup Patient Address Date Of Birth Ethnicity Gender Patient Name 3 p 20 Race 3j p 22 Age at Index Date G 0 3 EventDateTime it amp G 0 40 Disposition it G 0 41 Inpatient ICD9 Group 120 We now want to bring in APR DRG Left click on APR DRG and drag over to the Selected Attributes View Wide Long m Available Attributes cL A E a G 0 Common Attributes G 0 1 GroupNo k e G 0 1 GroupNo 1 3 D 1 PatientID v ES G 0 2 EventName F D 2 MRN m e G 0 3 EventDateTime 3 D 3 SES El G 0 4 EventID 3j p 14 Ethnicity a 5 G 0 5 EventType D D 15 Gender a E G 0 6 AdmitToInpatFromED Order Datetime E D 19 Preferred Language a 2 G 0 7 LengthOfStay Analytic days D D 20 Race Lc A G U 8 LengthOfStay days 3 D 22 Age at Index Date JA G 0 9 Admit Nursing Station A G 0 3 EventDateTime G 0 10 Admit Priority A G 0 40 Disposition G 0 11 Admit Referral Agency y G 0 41 Inpatient ICD9 Group A 6 0 12 Inpatient Admit Service Group iA 6 0 14 APR DRG Average LOS G 0 15 APR DRG Code APR DRG now appears in Selected Attributes x G 0 1 GroupNa e D1 PatientID 1 32 D 2 MRN FJ
116. x Study Designer TT reu W TTO Admit Ady TTO Admit A TTO Adv When In BasicWbcEchoabist X Reset Method Run Method View Wide gt Long s Available Attributes Selected Attributes F E 3 4 G 0 Common Attributes G 0 1 GroupNo E G 0 1 GroupNo D D 1 PatientID G 0 2 EventName M D 2 MRN G 0 3 EventDateTime 3 n 3 ses i o E A G 0 4 EventID D p 14 Ethnicity 3 G 0 5 tEventType 9 D 15 Gender l A 6 0 6 AdmitToInpatFromED Order Dateti 3 D 20 i Race E G 0 7 LengthOfStay Analytic days 33 D 22 Age at Index Date 4 G 0 8 LengthOfStay days A G 0 3 EventDateTime A G 0 9 Admit Nursing Station A G 0 40 Disposition G 0 10 Admit Priority G 0 41 Inpatient ICD9 Group G 0 11 Admit Referral Agency G 0 13 APR DRG Minimize the list populator SULLA St Run All saves e WPX X HS E Buh V Adv TTO Admit TTO Adv When In Reset Method Run Method Available Attributes Selected Attributes ethods Groups a E 3 4 6 0 Common Attributes E G 0 1 GroupNa i G 0 1 GroupNo TJ D 1 PatientID This results in 123 Study Designer p3 Run All E save As la ext x E je TTO Admit Adv TTO Admit TTO Adv When In BasicWbcEchoList X Reset Method Run Method us z o Outcome z ES 5 la End a Ne outcome Time from
117. y Summary 1 Data e 4 Now Save the Study Designer 114 List Method When you want to see attributes that were not used in the building of the original cohort Tutorial 23 List method listing attributes of events not involved in the cohort build https youtu be jhKGypDUU k In our previous discussion showed you how to obtain attributes belonging to the cohort s index event But suppose you wanted to look at other event types that were not involved in the cohort build Using our pneumonia patient example Suppose we wanted to see the first White Blood Cell count of the hospital admission or the first cardiac echo which occurred in the six months preceding the discharge Neither of these events were involved in building the cohort but they are of interest to us We want to use the machinery of the List Method to find these events and obtain specific values Our goal Produce from List Method in a single spreadsheet e Attributes from the cohort s index event o Disposition of the original pneumonia discharge Did they get discharged to a nursing home o The APR DRG of the original pneumonia These are two attributes of the original pneumonia hospitalization and are inherent to the event type e The First White cell Blood Count during the Admission This is an attribute of an event that was not used to build the cohort First White blood count during is a lab event that you are relating to the cohort through a When In
118. ys the parentheses under the estimates in the table do not overlap another indication of significance Left Click on the A Events tab E Qv WV Admission X Run Method Show Criteria 2 Demographics Target Event All Events Patient List a m Number Target B Events Achieved Person Days at Risk Incidence Density per year of risk Relative Risk Risk Difference g lt 124 135103 MES xe BadDiabetesWRepeatBadF2 456 366666 pie P o x Note Relative Risk group baseline group Relative Difference group baseline group The All Events Tab in the Simple mode counts all the admissions first second third after the index date By contrast the Target Event tab graph counts only the first admission to calculate the Cumulative percent of patients admitted at least once by time on the x axis The All events tab provides the relative risk of hospital admission considering all the admissions not just the first 42 Conclusion Diabetics with a bad repeat test are 1 355 times more likely to be admitted than Diabetics with a good repeat test 43 Tutorial 16 Importance of Cohort Index Date on the meaning of a Study Tutorial 16 Importance of Cohort Index Date on the Meaning of a study ppt6 https www youtube com watch v ovRRrYUj vA Objectives a Understand the impact of changing the Cohort Index Date on the nature of the question in study designer b Understand how

Study Designer V30 - Clinical Looking Glass from Montefiore

Contents

Download Pdf Manuals

Related Search

Related Contents