Company: Correctional Health Services-Los Angeles County Department of Health Services
Company background: Correctional Health Services (CHS) is the largest county-run correctional health system in the United States with a focus on social justice medicine, compassionate whole-person care, community partnerships, continuity of care, and professional growth.
Contact: Nazia Qureshi
Title: Epidemiologist
Country: USA
Award Category: Innovative Problem Solver
Tell us about the business problem you were trying to solve?
Our organization is responsible for healthcare delivery in a correctional (jail) patient population, which has unpredictable lengths of stay and a constant influx and outflux of people who are assigned different encounter-specific identifiers for each stay. Over the past year, we transitioned to a new electronic health record system and were concurrently handling three major public health crises-COVID-19, Monkeypox (Mpox), and opioid use disorder; therefore, we needed efficient systems in place to ensure the timely administration and tracking of multiple doses of COVID-19 vaccine, Mpox vaccine, and long-acting injectable buprenorphine (medication for opioid use disorder). Because vaccine and medication orders in the electronic health record are encounter specific and automatically discontinue upon patient release, we also needed the tracking system to account for individuals being released and identify them upon their return to the correctional setting with their most current front-end facing encounter-specific identifier to minimize breaks in medication and vaccine dose continuity.
How did you use SAS to solve that business problem? What products did you use and how did you use them?
In Base SAS 9.4 and SAS Studio 3.0, we combined data from our electronic health record system and utilized the transpose dataset by MERGE technique described in SAS technical paper 2785-2015. This entailed making use of a data step to create an identity number for each record using a back-end unique person identifier from the electronic health record, proc sql to create a macro variable containing the text needed to merge the files, and a final data step to merge the macro variable. This allowed us to transform multiple rows of COVID-19 and Mpox vaccine and injectable buprenorphine administration data into single rows with multiple columns for medication and vaccine administration ordered by date for each unique person. We also used conditional statements and date formatting to calculate and display future due dates for additional doses, proc sql to identify the current population due for additional doses, and count in proc sql to calculate the total number of doses administered, with the final dataset stored as sas7bdat files to serve as a permanent cumulative record.
What were the results or outcomes?
We were able to provide the correct number of COVID-19 and Mpox vaccine and injectable buprenorphine doses to our patient population using the correct dosing intervals, including individuals who were released and came back to our care under different identifiers, and maintain continuity with external vaccination records in the state immunization registry and records from our previous electronic health record system. We have administered over 40,000 COVID-19 vaccine doses, more than 1300 doses of Mpox vaccine, and more than 1900 doses of injectable buprenorphine.
Why is this approach innovative?
In a jail health system, it is challenging to rely on the electronic health record alone to schedule and complete vaccine and medication administration appointments due to the constant movement of patients between facilities, high turnover of patients in and out of jail, and staffing constraints. Combining various SAS techniques allowed us to create custom living registries and medication/vaccine appointment and administration tracking systems tailored to our unique needs. This facilitated efficiency and continuity of care, ensured accuracy in dosing intervals and dose types, and it ultimately helped us maximize the amount of COVID-19 and Mpox vaccines and injectable buprenorphine offered and provided to our patient population to prevent negative outcomes (i.e., spread of communicable disease, opiate overdose, and death). Data management and preparation through SAS is often associated with research, reporting, and surveillance, but it can also have an integral and active role in healthcare and continuity of care, especially when faced with a revolving door of patients.
/***multiple rows for medication administration per person***/
data MEDICATION_DATASET2;
set MEDICATION_DATASET;
by UNIQUE_ID;
if first.UNIQUE_ID then n=0;
n+1;
run;
proc sql noprint;
select distinct catt(' MEDICATION_DATASET2 (where=(n=',left(put(n,8.)),
') rename=(CLINSIGDATE=CLINSIGDATE',left(put(n,8.)),
' PRSNL_NM=PRSNL_NM',left(put(n,8.)),
' FACILITY=FACILITY',left(put(n,8.)),
' UNIT=UNIT',left(put(n,8.)),
' ROOM=ROOM',left(put(n,8.)),
' PRODUCT=PRODUCT',left(put(n,8.)),
' MANUF=MANUF',left(put(n,8.)),
' SITE=SITE',left(put(n,8.)),
' LOTNUM=LOTNUM',left(put(n,8.)),
' LOTEXP=LOTEXP',left(put(n,8.)),
'))')
into :mer separated by ' '
from MEDICATION_DATASET2
;
quit;
data MEDICATION_DATASET3;
merge &mer ;
by UNIQUE_ID;
drop n;
run;
/*Final solution has 1 row per person with multiple columns for multiple dates of vaccine or medication administration*/
