Abstract
The difficulty of obtaining accurate medication history from inmates at the time of incarceration is daunting. This article summarizes the success of a large urban jail in the use of online data to identify medication history upon incarceration. This article describes the scope of available prescription data, the implementation of online retrieval, system limitations, planned improvements, and suggestions of additional applications of online retrieval services.
Introduction
This article summarizes the experience of a large urban jail with online identification of prescriptions filled by inmates during the weeks preceding incarceration. It is difficult to obtain accurate and timely medication histories from inmates at the time of incarceration. The many difficulties have limited Philadelphia Prison System (PPS) telephone verifications primarily to cases of severe mental illness. Furthermore, if an inmate is taking but does not report taking medications, telephone verification is not initiated.
Verification online tripled the percentage of admissions for which PPS clinicians have credible prescription histories. This article reviews development to date of this service, its current use at PPS, its limitations, planned improvements, and likely future applications.
A Case Study
The chief medical officer of the PPS intake facility for males provided this example of the usefulness of access to complete prescription histories: This patient presented with a verbal history of seizures and schizophrenia. He claimed he was taking Depakote and Zyprexa, but he was unsure of his other medications. The intake nurse performed a medication history search, which returned a list of medications and the last time each prescription was filled. Although we remained unsure of the patient’s compliance prior to intake, the search provided insight into his concomitant medical problems. It revealed that he had been prescribed several benzodiazepines, a sedative hypnotic, an antiseizure medication, an antipsychotic, and an antihypertensive. This indicated the need to promptly refer to behavioral health, treat hypertension, and be on alert for withdrawal symptoms.
Background
PPS is the county jail for Philadelphia, Pennsylvania. The average fiscal year (FY) 2012 census was 8,800. Nearly 37,000 admissions to PPS occur annually. At the time of incarceration, 57% of PPS inmates reported that they had taken prescription medications prior to arrest.
Traditionally, PPS clinicians have attempted to verify inmate reports of current medication use by telephoning prescribers and pharmacies identified by inmates during the intake process. However, telephone verification is time consuming and largely unsuccessful for a number of reasons: Inmates often are unable or unwilling to provide accurate identification of prescribers and pharmacies; most pharmacies that might be called for histories are not open around the clock; telephone verification is time consuming; some pharmacists won’t provide information in the absence of written permission from the patient; often verification cannot be completed during intake evaluations; and in Philadelphia, the sheer number of pharmacies is large, exceeding 1,000. Thus, only 10% of PPS efforts to obtain prescription histories by phone produced useful data.
A Promising Technology
In 2011, PPS contracted with a prescription history verification vendor that offered electronic verification of prescription histories. This service works as follows: An electronic interface was created between the vendor and PPS’s inmate tracking system, which links all reports to the inmate’s unique police identification number, so that submission by medical staff of the police identification number results in the vendor also receiving all known inmate aliases. A link to the vendor’s secure website was installed on medical staff computers in each PPS intake facility. As part of the intake medical evaluation, and with the inmate’s written permission, a medical assistant enters the new arrival’s name, date of birth, and Social Security number into the vendor’s secure website. Within seconds, the vendor’s site shows a history of filled prescriptions for that inmate. These histories include the following data elements: inmate name/inmate home address/inmate gender/medication name/medication dose/frequency of administration/fill date/quantity/pharmacy name and address/prescriber name and address.
Intake clinical staff now uses online prescription history search results to elicit more complete treatment histories from inmates and to guide treatment planning. (In critical cases, if the online review shows no medications, staff will still call pharmacies or relevant providers to obtain histories).
Scope of Available Data
Filled prescription histories are retrieved from a commercial database that records the filling of prescriptions submitted electronically. Ninety-three percent of community pharmacies, including 85% of independent pharmacies, and 10 mail-order pharmacies accepted prescriptions electronically in FY 2012. Reportedly, this represents 69% of all office-based prescribers.
In Pennsylvania in 2012, about 46% of eligible prescriptions were submitted electronically versus 36% in 2011—almost a 30% increase in 1 year. This percentage is expected to continue to increase rapidly with continued adoption of electronic medical records technology.
Lessons Learned
Zip Code Issues: Originally, PPS searches were limited to pharmacies in the zip code of the home address provided by the inmates. The inaccuracy of home address reports and the frequency of patients’ use of pharmacies in zip codes other than their home zip code led us to change search criteria, so that all Philadelphia zip codes are searched for each inmate; doing so increased the percentage of inmates for whom histories were found from 6% to 25%.
Alias Issues: The average number of aliases per PPS inmate is 2.6. An interface was created between the inmate tracking system and the vendor’s system so that when an intake nurse submits the name provided by the inmate to the vendor, the vendor actually receives that name and all known aliases. However, this change did not materially increase the percentage of inmates for whom histories were found.
Data Volume Issues: Two-year histories often proved voluminous and not entirely useful. Searches are now limited to data from the 4 months preceding incarceration. A 4-month period was selected in order to capture data from more than one 90-day prescription period.
Access to antiretroviral treatment data is blocked by the data source.
The vendor arrangement described here provides current (active prescription) and recent (expired prescription) medication history data whereas the electronic medical record being implemented at PPS provides only current information.
Findings
The principal findings at PPS are that reliable prescription histories are now available for 3 times as many inmates as was achieved via the telephone process described previously, and prescription medication histories are available for the first time for at least some of those inmates who at intake report no prescription medication use. Four-month prescription histories are now found for 26% of male inmates and 40% of female inmates (by contrast, 49% of all male inmates and 73% of female inmates reported taking prescription medications in the community). Fifty-five percent of male inmate and 40% of female inmate reports matched online records entirely. Medication histories were found for 12% of those inmates who reported no medication history. Eighty-three percent of the inmates who reported taking medications reported fewer medications than reported online. Fourteen percent of inmates who reported taking medications reported more medications than reported online.
The above findings are presented with important caveats. First, the database is limited to prescriptions where prescribers sent prescriptions to pharmacies electronically; paper prescription data are not captured in this system. Also, the actual number of inmates who receive medications during the 4-month period preceding incarceration is not known.
Additional Uses
PPS has been working to measure continuity of care issues after discharge. In a recent PPS study, a sample of inmates medicated at PPS for seizures, HIV, or diabetes were asked on reincarceration for permission to look at filled prescription data for the period between release and reincarceration. HIV-related inquiries were eliminated, given that most antiretroviral drug information is blocked by the data source. Of the 188 individuals remaining in the sample, only 11 received medications between incarcerations. This suggests that a large percentage of releasees discontinued care upon release; alternative explanations may be plausible. In the future, medication histories might be used to identify a variety of continuity-of-care challenges.
Implications for Clinicians
It is now possible to better identify, diagnose, and plan treatment for medically compromised inmates by introducing online recovery of medication histories at the time of intake. Filled prescription information can be integrated into electronic medical records. Increased use of electronic medical records systems should enlarge the database quickly. The database can contribute to monitoring of continuity of care.
We speculate that other providers, such as emergency room clinicians, may also benefit from access to filled prescription databases.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. For information about JCHC's disclosure policy, please see the Self-Study Exam.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.