Electronic medical record reduces HIV medication refill response time and emergency refills in a Latino community clinic

INTRODUCTION

Strict adherence to the antiretroviral regimen in highly active antiretroviral therapy patients is essential for effective treatment. ^1–3 It is therefore imperative that providers seek to decrease opportunities for missed doses due to health-care system deficiencies. System deficiencies can be present in many areas of the health-care system, including the systems that handle prescriptions and refills.

A step that has not been concentrated on much in the past is the transmission of prescriptions, by various means, to the pharmacy.

Utilization of facsimile or phone transmission for medication refill requests allows multiple opportunities for potential error and delayed approval of the processing of refill requests through poor verbal communication, poor handwriting or poor resolution of faxes. Faxes or phone calls can be accidentally made to the incorrect pharmacy. Or, when a provider asks a staff member to fax or call the prescription in, this adds another step for potential human error.

It is well known that recently Medicare is offering a 2% financial incentive for a year for those practices that can electronically prescribe, since it is time- and cost-effective and increases quality of care. ⁴

The clinic had delays in processing medication refill requests to the point that HIV patients were at risk of missing doses – a critical error causing potentially irreversible damage. In order to compensate for this discrepancy, the serving pharmacies issued five-day emergency antiretroviral refills when they did not receive an answer from the clinic for the refill request.

HIV is a chronic infection that can be well-controlled by chronic medication administration. Resistance arises when medication levels are subtherapeutic due to missed doses. These resistance mutations are archived for the lifetime of the patient, causing resistance to the original medicine, to other related medications, and perhaps to co-administered antivirals if not dealt with promptly. It is far more advantageous to prevent resistance than to discover it and manage it over the rest of the patient's life. Resistance may mean that the patient now must take medications that are potentially more toxic than the previous regimen, causing significant morbidity and potentially less adherence to the new regimen – a vicious downward spiral to possibly serious morbidity and eventual mortality.

In May 2006, the clinic launched an electronic medical record (EMR) system, which streamlined the process of prescription refill approval.

The goal in conducting this investigation is to determine whether the EMR system does, in fact, result in a reduction of the emergency refills issued and a reduction in the overall time required for a prescription refill to be approved.

METHODS

This is a retrospective data analysis of two four-week periods evaluating the efficiency of using manual methods of communication between three HIV clinics and two private pharmacies versus an EMR system. The number of emergency refills and the time to respond to the pharmacy-generated refill requests are the two variables measured.

We surveyed the multiple pharmacists in the two pharmacies for the number of emergency refills per month. They used their data and recall of staff to generate these numbers. To compare the refill response time and number of emergency refills, we also surveyed the pharmacists for their opinion comparing other HIV clinics with a similar prescription volume for these functions.

We also analysed pharmacy and clinic records to determine the number of days passed between the request and the communication back to the pharmacy. Specific data from the pharmacy software were used to determine the response time to refill requests along with handwritten notations and fax copies.

Multiple medications requested for the same patient on the same day were treated as one data point.

AltaMed has three HIV clinics in East Los Angeles serving ∼1400 mainly underinsured and Latino patients. There were four providers prescribing medication in 2006 and three in 2007.

Two locations of MOMS Pharmacy, an HIV specialty pharmacy (www.momspharmacy.com), provided both hard-copy data and electronic pharmacy record data for one-month periods in 2006 and 2007, not counting weekends and holidays.

HealthMatics EMR (http://www.a4healthsystems.com/) and the web-based service SureScripts (http://www.surescripts.com/) were implemented in May 2006. SureScripts is a service that provides an interface between EMRs and pharmacy software so that what is written in the EMR is transmitted into the pharmacy software in about 5 minutes; the pharmacist must then make minimal keystrokes to print a label. There is no transcription of the data.

The number of emergency refill requests and response times for each prescription were obtained for the one-month time period of 13 February to 12 March in both years. Using the same period in a year controls somewhat for seasonal variations in prescription refill requests. It also gave enough time for the EMR to be adapted to by the clinic and pharmacies. We compared data from before the EMR start date and after 10 months of use: the number of emergency refills from two specific 28-day periods and the response times of the clinic to refill requests measured in days from zero (same day) upwards.

Requests with no request date or requests with no refill date were assigned question marks and tallied separately. These were not included in the analysis due to unknown response times; these represent 4.82% of the refills.

Data were analysed utilizing Microsoft Excel 2003. Average refill times for each pharmacy as well as a combined average were calculated. To determine statistical significance of the results, a two-sample t-test assuming unequal variances (heteroscedastic t-test) was also calculated. Variance was calculated utilizing a two-sample F-test for variances.

RESULTS

The number of emergency refills decreased at the two pharmacies from 66 and 12 to less than one per month (Table 1). There was also a subjective improvement of the refill process compared with other HIV clinics served by these pharmacies.

The average time to a response to a pharmacy refill request decreased from 1.57 days to 1.04 days (P < 0.004; Figure 1). Data here for both pharmacies were combined for this measurement since we did not see a significant difference in the inter-pharmacy data. The variance of the curve was 3.53, dropping to 1.73 with EMR. Before EMR, the maximum response times were nearly three days longer than with EMR.

OPEN IN VIEWER

Figure 1

Comparison of pre- and post-electronic medical record (EMR) refill response times

DISCUSSION

The results of this study demonstrate a 0.5-day decrease (P < 0.004) in refill approval time as well as a decreased variance. This can in turn lead to a decreased opportunity for missed medication doses by ensuring a patient receives his/her medication refill on time.

Pharmacy perception of improved refill request response time and decreased emergency refill issuance after EMR and compared with the same clinics and other clinics (accounting for clinic prescription volume) also demonstrates increased efficiency with the utilization of EMR compared with facsimile or phone communication.

These data do not identify, but imply, a possible positive correlation between EMR utilization and patient medication adherence by improving one step in the system.

These results have the following limitations: the clinic had four providers in 2006 and three in 2007; 16 data points (4.82%) were not included in the statistical analysis due to unknown refill approval time.

The survey of pharmacists is subjective data, but is relevant to this study because it is indicative of the value of EMR relative to the relationship between the clinics and pharmacies.

Further research should be done with larger sample sizes in order to confirm these results.

In addition, research should be done to see whether the results of this study apply to the clinical setting outside of HIV medicine. Previous research has demonstrated that EMR increases efficiency when compared with handwritten methods in a setting of chronic illness and multiple medications, which would imply that EMR has increased application to the specialty of HIV medicine. ⁴

EMR is used to address many problems in clinics. Simbini ⁵ points out that the complex management of HIV patients is better managed by an EMR. Some of these efforts have focused on loss-to-follow-up data ⁶ and drug interactions or incorrect dosing of complex medication regimens. ⁷ There have been multiple studies of EMR implementation for HIV care. There are descriptions of new implementations for the Philippines, ⁸ Rwanda ⁹ and Kenya. ¹⁰ These have focused on in-house pharmacies, ¹¹ not on independent pharmacies located outside of the clinics as this present study does.