ACADEMIC EMERGENCY MEDICINE 2012; 19:102–105 © 2011 by the Society for Academic Emergency Medicine
Objectives: The primary objectives were to assess whether electronically delivered prescriptions lead to reduced pharmacy wait time, improved patient satisfaction, and improved compliance with prescriptions. Secondary objectives included determining other reasons for noncompliance and if there was an association between prescription noncompliance and subsequent physician and emergency department (ED) visits.
Methods: In this prospective study, patients discharged from the Ronald Reagan UCLA Medical Center ED with prescriptions for nonnarcotic medications were randomized to a control group who were discharged with standard written prescriptions or an intervention group who had their prescriptions electronically delivered to the pharmacy of their choice. All study participants were contacted 7 to 31 days after ED discharge for a structured telephone interview.
Results: Of the 454 patients enrolled, follow-up was successful for 224 patients (52.4%). Twenty-eight patients did not fill their prescriptions (12.5% noncompliance rate). The top three reasons patients stated for not picking up their medications were perceiving their prescription as unnecessary (n = 11), medication affordability (n = 5), and lack of time (n = 4). There was no difference in primary prescription noncompliance between the two study groups (p = 0.58). However, electronically delivered prescriptions significantly reduced the median pharmacy wait time, from 15 to 0 minutes (p < 0.001), and improved patient satisfaction at the pharmacy (p = 0.034). Neither subsequent physician nor ED visits were increased by primary prescription noncompliance.
Conclusions: Electronically delivered prescriptions significantly minimized pharmacy wait time and improved patient satisfaction at the pharmacy, but did not improve primary compliance with prescriptions.
Primary prescription noncompliance is potentially a significant issue among discharged emergency department (ED) patients. Previous studies have documented primary prescription noncompliance in 12% to 25.8% of discharged patients1–4 and as high as 35% in one study targeting pediatric discharges.5
Research in primary prescription noncompliance has generally been limited to chronic diseases in the outpatient setting.4 Only a handful of studies have targeted ED patients. Previous research has reported several reasons for prescription noncompliance, including medication affordability, long pharmacy wait times, medication concerns, perception of medication as unnecessary, and lack of time.6,7 Research on strategies to improve primary prescription compliance among discharged ED patients is lacking.
The primary objectives of this study were to assess whether electronically delivered prescriptions lead to reduced pharmacy wait time, improved patient satisfaction, and improved compliance with prescriptions. Secondary objectives included determining other reasons for noncompliance and if there was an association between prescription noncompliance and subsequent physician and ED visits.
This was a prospective, randomized study. The UCLA institutional review board approved the study protocol. Informed consent was collected from all participants.
Study Setting and Population
The study was conducted at the Ronald Reagan UCLA Medical Center ED, an academic medical center ED with an annual census of 45,000 visits. All patients discharged from the ED with a prescription for nonnarcotic medications were eligible for the study. Exclusion criteria included hospital admission, psychiatric complaints, lack of contact information, inability to perform a follow-up telephone interview, over-the-counter medications, schedule 1 or 2 narcotic prescriptions, inability to understand the English consent forms, and non-California residence. We also excluded patients who delayed picking up their medication from the pharmacy, to prevent noninterpretable data.
Research associates identified all eligible study participants by monitoring the ED information system for patients discharged with prescriptions, reviewing patient discharge notes, checking prescription printers, and querying nurses and/or physicians. Research associates enrolled study participants from 8 am to 12 am, 7 days a week, from June 22, 2010, to December 25, 2010. Research associates provided a description of the research protocol to eligible patients and obtained informed consent for study enrollment. Study investigators, research associates, and study participants were blinded to study arm enrollment until after consent was obtained. Study participants were randomized into a study arm (control group or intervention group), using patient assignment by random number with assignment group in sealed envelopes. Patients in the control group received no alteration to care and were discharged with the standard prescription that was to be filled on their own. Patients in the intervention group received a prescription that was also electronically delivered via facsimile directly to the pharmacy of their choice. Demographic information on patient sex, age, and level of education were also obtained.
Study outcomes were collected by performing telephone interviews after ED discharge. All data were self-reported. Research assistants attempted to contact each study participant for a structured telephone interview 7 days after ED discharge. However, for patients who could not be reached, continued contact was attempted up to 31 days after the ED visit. Every patient was asked the same series of questions that included the following, in this order: 1) whether medications were obtained from the pharmacy; 2) number of prescriptions filled compared to the number prescribed; 3) reason for not obtaining medications (medication cost, perceiving medication as unnecessary, fear of addiction or side effects, long pharmacy wait time, inconvenience/lack of transportation/lack of time, sufficient supply of medication, lost prescription, or other); 4) wait time estimated from the time the patient reached the pharmacy counter to the time medications were dispensed; 5) rating of patient satisfaction with pharmacy service (poor, fair, good, very good, or excellent); and 6) number of subsequent ED visits and physician visits in the 7 to 31 days following ED discharge.
Data analysis was performed with STATA 9 (StataCorp, College Station, TX), using the Fisher’s exact and chi-square tests for categorical variables and the Wilcoxon rank-sum test for nonnormally distributed continuous variables. Although there were several components to the primary outcome, we chose not to make adjustments for family-wise Type I error rates. Initial sample size calculations were done with STATA 9. Based on previous studies in non-ED settings, we assumed that 80% of ED patients fill their prescription when given a prescription and that 90% fill them when arrangements are made for their medication to be ready for pick-up by faxing the prescription to the pharmacy. Using these estimates, a sample size of 137 in each group was required for a power of 0.8 and an alpha of 0.05.
During the study period we assessed for eligibility 1,114 ED patients discharged with a prescription. A total of 647 did not meet inclusion criteria or declined to participate. Of the remaining 467 patients, 239 were randomized into the control group and 228 into the intervention group. Thirteen were later excluded due to exclusion criteria. Of the remaining 454 patients contacted for the telephone interview, 211 were lost to follow-up and five withdrew (52.4% successful follow-up). Fourteen patients who delayed picking up their medications were later excluded per our exclusion criteria. Data analysis was performed on the remaining 224 patients.
The subjects’ median age was 32 years, with an interquartile range (IQR) of 6 to 85 years. Females accounted for 60.3% of all patients. Demographic distribution (age, sex, and level of education) was similar among the standard and electronically delivered prescription groups.
Of the 224 patients, 196 went to the pharmacy and picked up their medication, yielding a primary compliance rate of 87.5%. Of the 117 subjects with electronically delivered prescriptions, 101 (86.3%) picked up their medications, and of the 107 standard prescription subjects, 95 (88.8%) picked up their medications. These primary compliance rates were not significantly different (p = 0.578).
Of the 28 patients who did not fill their prescriptions, 11 (39.3%) felt that their prescription was unnecessary, five (17.9%) could not afford their medications, four (14.3%) lacked the time, three (10.7%) feared medication side effects, two (7.1%) already had a sufficient supply at home, one (3.6%) was physically incapable of going, and two stated other reasons. Among the 11 who thought their prescription was unnecessary, seven were discharged with hydrocodone/acetaminophen. However, overall 165 of 172 (95.9%) patients who were prescribed hydrocodone/acetaminophen filled this prescription.
To assess whether faxing prescriptions yielded a shorter pharmacy wait time, patients were asked if medications were ready for immediate pick-up. Immediate pick-up was defined as medications that were available for pick up upon reaching the pharmacy counter. It was made clear to patients not to include time spent traveling to the pharmacy and time spent in line. Medications were ready for immediate pickup for 53.5% (n = 54) of the 101 electronically delivered prescription patients versus 7.5% (n = 7) of the 93 standard prescription patients (p < 0.001). The median pharmacy wait time was 0 minutes (IQR = 0 to 15 minutes) for patients with an electronically delivered prescription and was 15 minutes (IQR = 10 to 25 minutes) for patients with a standard prescription (p < 0.001).
Electronically delivered prescriptions resulted in greater patient satisfaction with the process of obtaining medications: 66.3% of patients in the electronically delivered group rated their satisfaction with the process of obtaining medications as very good/excellent, compared to 48.4% in the control group (p = 0.034). There was no significant difference in the rates of return ED visits (p = 0.480), or subsequent physician visits (p = 0.152), between patients who filled their prescriptions and those who did not.
The rate of primary prescription noncompliance (12.5%) we found is similar to the 12% reported by Thomas et al. in 1996.1 The majority of patients who did not fill their prescriptions felt their medications were unnecessary and would provide little benefit (39.3%). Of these, seven were prescribed hydrocodone/acetaminophen. Past literature shows that patient perception of medication benefit correlates to prescription compliance.7
Cost was the second most common reason for primary prescription noncompliance (17.9%). Three previous ED studies have found cost to be a deterrent for 10.3, 36, and 50% of patients, and it remains consistently as one of the top three reasons for patients not filling their prescriptions.2,4,7,8
Lack of time was the third most common reason for prescription noncompliance (14.3%). Campbell et al.9 reported 9 of 42 patients (21.4%) stating “busy” as a reason for not picking up their medications. Matsui et al.10 documented 6.8% of parents of pediatric ED patients not filling prescriptions due to lack of time. Moreover, four studies targeting non-ED populations that found lack of time as a contributor reported a range of 0.07% to 6.2%.6,11–13 These studies suggest the need for interventions that minimize pharmacy wait time.
To our knowledge, this study is the first to assess the effects of an electronically delivered prescription intervention on primary prescription noncompliance among ED patients, and we did not discover a difference in prescription fulfillment rates between standard and electronically delivered prescriptions. Potential reasons for this are 1) ED patients generally suffer from emergent illnesses and, therefore, will be likely to fill their prescriptions regardless of the prescription delivery method, and 2) this study was conducted at an ED that primarily services a geographic area with 74.8% of patients reporting an annual income of $30,000 or more and 80% of patients with health insurance.14 The study population may not reflect the general U.S. population.
Although we did not find electronic delivery of prescriptions to improve primary prescription compliance, our intervention did achieve a significant reduction in pharmacy wait time. The median wait time was reduced from 15 minutes to 0 minutes. While none of the patients contacted for the follow-up interview stated long pharmacy wait times as a deterrent for primary prescription compliance, previous studies have shown long waits at the pharmacy to contribute to noncompliance.7,15
With the growing use of electronic medical records in the ED, e-prescribing will become more widespread. We anticipate the results of e-prescribing systems, such as the recent Centers for Medicaid and Medicare Services E-Prescribing Incentive Program, to result in reduced pharmacy wait time and greater patient satisfaction.
One essential point that should be mentioned is the effect of e-prescribing on pharmacy and overall health care costs. The pharmacist fills these prescriptions in advance and must “unfill” them if they are not picked up, which causes extra work and increases the risk of medical error associated with restocking medications.
Study outcomes were based on patient self-report. Patients were contacted for a telephone questionnaire 7 to 31 days after ED discharge. The time between the pharmacy visit and the telephone questionnaire may have allowed for recall bias. We did not compare our compliance rates with actual insurance data. Hohl et al.2 found that patients overreported fulfilling their prescription by 7%. Research assistants were not blinded to the study group during the telephone interview, which may have resulted in observer bias. In the electronically delivered group, 19 patients encountered faxing problems that included nonfunctional fax machines, pharmacies not accepting faxed prescriptions, and transmission errors. These problematic issues associated with electronically delivered prescriptions contributed to delayed medication retrieval from the pharmacy and consequently poorer ratings of pharmacy services. Our study was performed in a single ED that generally services a wealthier, insured population. Therefore, our results may suffer from lack of external validity. Finally, the low rate (52.4%) of successful follow-up may also decrease the generalization of our findings.
Electronically delivered prescriptions did not result in decreased primary prescription noncompliance; however, they did minimize pharmacy wait time and increased patient satisfaction at the pharmacy.
The authors thank Guy Merchant for his invaluable assistance in creating the study database and aid in completing the study. We also thank the Emergency Medicine Research Associates for their immense help with patient enrollment.