GPC Network



PVRM 868: Medical Informatics Driven Clinical Research Fall 2016Cohort Development Examination: CDExamFall2016This is a take home exam. You will need access to HERON to complete this assignment. Please conduct this assignment independent of your other classmates. This exam will test your ability to properly construct queries using i2b2, its features, but also acclimate to the kinds of data available in HERON.Most questions have only one correct answer; the number returned by i2b2 for the count size of the cohort. In general, unless otherwise indicated, use the de-identified dates as they are currently existing in HERON when defining cohorts. Also, consider using the timeline to help you visualize the cohorts you are defining. It can really be your friend as these questions get more complex regarding same financial encounter and temporal queries.Output:In your personal folder, create a folder “PVRM868CDExamFAll2016. In it you will place your queries for the questions. Please label them “Q1_addedtextthathelpsyourrember”, “Q2…” Also send back this exam word document with the number as the answer for each questionYou may also include further narrative description justifying your answer in this document as well. Some questions are so straightforward it’s not necessary but for the latter questions this may be advisable to obtain partial credit. Due date: please email your document to Russ and Tamara and be sure to save your queries by next Wednesday, September 28, 2016. Questions:Demographics: How many patients are identified as being “female” and having “black or African American” as their race. Cancer: Of those, how many have breast cancer as defined by the Tumor Registry, SEER Site SummaryOf those, how many had the assignment of breast cancer defined by SEER Site Summary since 1/1/2010.Family History: You have heard that Epic captures “family history” (Hint: look in the history folder). How many of these women from Q3 have family history of breast cancer recorded? Relationships: But how many of those have a positive history of breast cancer in their mother or maternal grandmother? Diagnoses: Create a cohort for Myocardial Infarction based upon the “most commonly occurring” ICD9 code category (3 digit) and also include the most common ICD10 code category (3 character)Please also indicate in this document which 3 digit ICD9 code you used and which 3 character ICD10 code you used: How many patients do you have with diagnoses covered by using either that ICD9 code or the ICD10 code: Diagnoses modifiers: now using only the ICD9 coded patients using that single 3 digit code, how many have had that diagnosis code category observed in any part of O2 (the EMR): Diagnoses Modifiers: For the Q7 cohort, how may have had that same ICD9 code recorded in the University Healthsystem Consortium (Hospital diagnosis UHC) source system at any time: Same Financial Encounter: For Q8, how many have had both the O2 and the UHC assigned coding happen during the same financial encounter: UHC: Another way to look at inpatients with myocardial infarction could be to use the “AHRQ Quality” metric (Agency for Healthcare Quality and Research) that evaluates inpatient mortality for cases with AMI (acute myocardial infarction). What’s the intersection of patients from Q9 with the In-hospital mort AMI criteria during the same visit? UHC isolated and medications: Now, just focus on using the AMI population defined by the AHRQ measure (1,879 patients). For those patients how have ever been exposed to an ACE Inhibitor? Medications modifiers: how many had an outpatient prescription for an ACEI during the hospitalization where the AHRQ AMI quality measure used in #10 occurred? Medications, temporal queries and trial recruitment: You are interested in pursuing a funding opportunity from AHRQ regarding medication prescribing and adherence of ACE Inhibitors. You want to do a prospective study to survey AMI patients regarding their AMI hospitalization and if they understand the importance of taking ACE Inhibitors. Using the AHRQ AMI population (2,001 facts; 1,879 patients). You want to survey people who have had:Only one hospitalization that met AHRQ criteria (eliminate those who had 2 or more hospital admissions that met AHRQ AMI criteria): For those, then exclude the number of people that may be deceased. What is the resulting cohort size? Building on Q13b, you want to use this cohort to ask patients who were ACEI na?ve (wrt just KUMC prescribing) prior to their admission questions about how they take their medication. Build a temporal query that focused on the Q13b cohort who have only had an ACE Inhibitor prescription (outpatient order) during or up to one year after their hospitalization where they were denoted as meeting the AHRQ AMI criteria.Rerun Q14 but save the patient list. Now use that patient list and determineUse the “Patient Set” from this patient list query and now determine how many of them have used MyChart? You want to do your survey electronically to save on postage: Hmm… the cohort size from Q15 may be small if you get a poor survey response. How many more can you reach if you also include patients who have an email on file in the Epic EMR? You have heard there is program through the Frontiers CTSA program where you can contact patients without having a treating relationship with them “Frontiers Research Participant Registry”. How many patients in Q16 might you reach through the Frontiers Research Participant Registry? This question requires a written answer but if you want you can also support your argument with additional queries. Based upon Q17, would you proceed with the survey based only upon the Frontiers Participant Registry population, or would you seek to engage clinicians with treating relationships for conducting the survey? Justify your approach.Assuming you proceeded with broader outreach, what might be a strategy to identify the best clinics or clinicians to engage in support of your survey?Bonus points: do question 14 but truly define patients who you think are ACEI na?ve and not just KUMC prescribing na?ve prior to their hospitalization ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download