“Meds-in-Hand” Intervention to Reduce Critical Process Delays in Pediatric Human Immunodeficiency Virus Post-Exposure Prophylaxis

Abstract

Pediatric human immunodeficiency virus post-exposure prophylaxis is frequently indicated, but delays in medication receipt are common. Using plan-do-study-act cycles, we developed a multidisciplinary collaboration to reduce critical process delays in our pediatric emergency department. Interruptions decreased from a median 1 per month pre-intervention to zero per month during the intervention.

Approximately 40%–50% of youth who present to pediatric emergency departments (EDs) after sexual assault are eligible for human immunodeficiency virus (HIV) post-exposure prophylaxis (PEP) [1]. Current Centers for Disease Control and Prevention guidelines recommend 3-drug antiretroviral PEP to start within 72 (ideally 24) hours after exposure to potentially HIV-infected blood or bodily fluid, to continue for 28 days [2, 3]. However, studies in adults and children consistently demonstrate poor prescription fill rates, prescription errors, and obstacles to PEP adherence [2, 4–8]. Challenges unique to pediatric PEP include that most pharmacies do not routinely stock pediatric (liquid or chewable) antiretroviral formulations and some insurance policies require prior authorization for these medications.

To mitigate delays, starter packs are standard practice for ED patients prescribed HIV PEP [2]. Starter packs include a 3- to 7-day medication supply given to a patient until the full 28-day prescription can be obtained. A South African study, however, demonstrated a 3-fold increase in regimen completion with dispensation of a full 28-day regimen compared with starter pack provision alone and a plan for short-term follow-up for the remaining prescription [9, 10].

Boston Medical Center (BMC), a large safety-net hospital, routinely provides HIV PEP starter packs and infectious diseases consultation and follow-up. However, delays in PEP receipt occurred frequently. A previous study conducted at BMC demonstrated decreases in asthma readmissions with prescriptions filled and delivered by pharmacy to asthma patients prior to discharge [11]. We aimed to replicate this model to provide 28-day HIV PEP delivery in the pediatric ED.

Under the framework of the Model for Improvement [12], we developed 3 plan-do-study-act (PDSA) cycles with aims to initiate and evaluate a “Meds-in-Hand” program, in which a pharmacy representative delivers a 28-day filled prescription to an ED patient during the index visit, and to reduce to zero the number of patients with an HIV PEP interruption.

METHODS

In May 2016, we formed an HIV PEP working group that encompassed personnel from inpatient and outpatient pharmacies, pediatric infectious diseases, and pediatric emergency medicine to evaluate and improve HIV PEP medication receipt. Here, we describe process measures introduced from June 2016–October 2016 and continued to date.

Intervention

In June 2016, we initiated Meds-in-Hand in the pediatric ED (PDSA cycle 1). After an HIV PEP-eligible patient is identified, an ED clinician consults the pediatric infectious diseases department, contacts the ED pharmacy (24-hour coverage) to initiate pharmacy workflow, and orders an immediate ED PEP dose plus either a 3-day starter pack and 24-day prescription (if after outpatient pharmacy business hours) or a 27-day prescription for Meds-in-Hand (if during outpatient pharmacy business hours). The outpatient pharmacy, if open during the index ED visit, fills prescriptions and delivers Meds-in-Hand to the patient before discharge. When the outpatient pharmacy is closed, the inpatient pharmacy provides the starter pack, and the ED pharmacist ensures the prescription is appropriately faxed or e-prescribed to an outpatient pharmacy to be filled during business hours.

In August 2016, we added PEP workgroup e-mail communications, triggered by the on-call pharmacy manager (PDSA cycle 2). These e-mails detail Meds-in-Hand or starter pack provision and the pharmacy receiving the prescription(s). Follow-up e-mails document any delays or successful patient prescription fulfillment.

Initial process measure evaluation prompted informed development of an algorithm (Supplementary Figure 1) to guide pediatric ED clinicians in timely and appropriate pharmacy workflow use, pediatric infectious diseases consultation, medication selection, baseline laboratory testing (including hepatitis B surface antibody and antigen), and empiric gonorrhea, chlamydia, and trichomonas treatment as applicable [2, 3]. In October 2016 (PDSA cycle 3), we developed a mechanism to provide current and new ED staff this algorithm.

Study of the Intervention

We analyzed electronic medical charts and PEP e-mails from pediatric ED patients prescribed HIV PEP from January 2016–May 2016 (pre-intervention period) and June 2016–December 2017 (intervention period). Eligible patients were identified retrospectively via pediatric infectious diseases consult lists and prospectively from pharmacy-initiated PEP e-mail communications. We abstracted demographic characteristics, exposure type, medications prescribed, follow-up HIV testing, and course completion. The pharmacy department provided prescription and starter pack fill dates and times.

Measures and Analyses

We defined HIV PEP interruption as ≥24 hours elapsing for daily regimens or ≥12 hours for twice-daily regimens between ED medication dose, or last projected starter pack dose if given, and full prescription receipt. Descriptive statistics, t tests for continuous variables, and Fisher exact tests for categorical variables were used to characterize the population. Using run charts, each month we measured the median number of patients who experienced a PEP interruption, defined as preventable if caused by an insurance, pharmacy stock, or prescriber error and as nonpreventable if due to delayed pick-up. We also measured the median number of patients who received either a starter pack or Meds-in-Hand. We adjusted the median (center line) at the occurrence of any instances of special change variation [13].

Balancing Measures

As a balancing measure, we documented time from ED presentation until first HIV PEP dose administered to determine whether the additional tasks required by the new workflow delayed this critical step. We compared this measure using an exact Wilcoxon rank sum test given the small sample size. To evaluate potential obstacles, an online questionnaire about PEP algorithm awareness and use was e-mailed to pediatric ED physicians during summer 2017.

This study, implemented as a quality improvement project, did not require institutional review board approval per institutional guidance.

RESULTS

From January 2016–December 2017, 96 pediatric ED patients were evaluated for blood or bodily fluid exposures. Twenty-one patients did not meet HIV PEP administration criteria, that is, no clear exposure occurred (n = 10), presentation ≥72 hours post-exposure (n = 10), or declined treatment (n = 1). Seventy-five patients were prescribed HIV PEP. Age, sex, exposure type, and medication regimens did not differ between pre-intervention and intervention periods (Table 1).

Table 1.

Demographic, Exposure, and Post-Exposure Prophylaxis (PEP) Regimen Data for Individuals Prescribed Human Immunodeficiency Virus PEP During the Study Period

Characteristic	All, N = 75 (%)	Pre-Intervention, N = 11 (%)	Intervention Period, N = 64 (%)	P Valuea
Mean age, y (range)	15.3 (1–22)	12.9 (1–20)	15.7 (2–22)	.11
Female	61 (81)	10 (91)	51 (80)	.68
Type of exposure
Sexual assault	56 (75)	7 (64)	49 (78)	.47
Needlestick	10 (13)	2 (18)	8 (13)
Other	8 (11)	2 (18)	6 (10)
Time period seen
Weekdayb	23 (31)	5 (45)	18 (28)	.33
Weeknight	27 (36)	2 (18)	25 (39)
Weekend	25 (33)	4 (36)	21 (33)
PEP regimen prescribed
TDF/FTC + Dolutegravirc	59 (79)	7 (64)	52 (81)	.14
TDF/FTC + Raltegravir	2 (7)	1 (9)	1 (2)
Zidovudine, Lamivudine + Raltegravird	14 (19)	3 (27)	11 (17)
Dosage form
Liquid/Chewabled	14 (19)	3 (27)	11 (17)	.42
Pill	61 (81)	8 (73)	53 (83)
PEP delay
No delay	47 (63)	2 (18)	45 (70)	.003
Delay	13 (17)	5 (45)	8 (12)
“Preventable” delaye	7 (9)	5 (45)	2 (3)
“Nonpreventable” delaye	6 (8)	0	6 (9)
No data	15 (20)	4 (36)	11 (17)
Follow-up
Stopped PEP once source tested negative	2 (3)	0	2 (3)	…
Completed course
Yes	26 (36)	5 (45)	21 (34)	.37
No	8 (11)	2 (18)	6 (10)
Unknown	39 (53)	4 (36)	35 (56)

Abbreviations: PEP, post-exposure prophylaxis; TDF/FTC, tenofovir disoproxil fumarate/emtricitabine.

^a P values represent Fisher exact test for categorical variables and t test for continuous variables comparing pre- and during-intervention period characteristics.

^bWeekday: 8 am–7 pm Monday–Friday (pharmacy hours); weeknight: 7 pm–8 am Monday–Thursday; weekend: 7 pm Friday–8 am Monday.

^cThis study occurred prior to the May 2018 Centers for Disease Control and Prevention and World Health Organization recommendations to avoid dolutegravir for human immunodeficiency virus PEP in women who could possibly be less than 8 weeks pregnant or become pregnant while taking PEP [2]. Institutional guidelines changed at that time to reflect the update, substituting raltegravir for dolutegravir in applicable patients.

^dOne young child was prescribed liquid zidovudine, lamivudine, and lopinavir/ritonavir initially. Due to side effects (nausea/vomiting), the lopinavir/ritonavir was switched to raltegravir dissolvable granules; 2 children received lamivudine/zidovudine combination tablets and chewable raltegravir tablets; all other children in this category received zidovudine and lamivudine liquid plus chewable raltegravir tablets.

^e“Preventable” delays refer to prescriber-, pharmacy-. or insurance-related issues, including prescription errors. “Nonpreventable” delays refer to delayed patient pick-up despite filled prescriptions and attempts made to contact the patient.

PEP interruptions decreased from a median 1 interruption each month pre-intervention (45% of PEP prescriptions; Figure 1) to a median zero interruptions during intervention months (13% of prescriptions; median shift, with 6 data points below the previous median). Whereas prescriber, pharmacy, or insurance issues caused all pre-intervention delays, most intervention-period PEP interruptions were secondary to “nonpreventable” patient pick-up delays. All patients prescribed liquid/chewable formulations pre-intervention and none during the intervention had PEP delays. We could not measure delays for 4 (36%) individuals before and 11 (17%) during the intervention because prescriptions were not filled at a BMC pharmacy.

Figure 1.

Run chart: human immunodeficiency virus post-exposure prophylaxis (PEP) prescription delays. Bars indicate the number of patients each month prescribed PEP; orange indicates number with delay in medication receipt causing PEP interruption (lapse ≥24 hours for daily regimens, or ≥12 hours for twice daily regimens, between first PEP dose and any subsequent doses); blue indicates those without an interruption; and gray indicates those with insufficient data to determine. Solid line indicates number of patients each month leaving without a starter pack. Dashed line represents median number of patients with an interruption or who failed to receive a starter pack (lines overlap as median shift occurred in same month). Arrows indicate PDSA cycles. Abbreviation: PDSA, plan-do-study-act.

In 3 of 4 pre-intervention months, ≥1 patient left without a starter pack or Meds-in-Hand. This failure occurred during only 1/19 intervention months, creating a median line shift (Figure 1). Seventeen of 21 (81%) intervention-period patients evaluated during outpatient pharmacy business hours received Meds-in-Hand (1 patient declined; 3 algorithm failures occurred).

Of 34 patients with BMC follow-up, 26 (76%) completed PEP (Table 1); overall, 57% prescribed liquid/chewable formulations and only 27% prescribed pills completed PEP. Due to missed visits or follow-up elsewhere, 39 patients (53%) had unknown PEP completion. Forty-one (55%) had documented follow-up HIV testing (all negative; data not shown). The pharmacy assisted 17 patients by facilitating insurance enrollment, co-pay discounts, medication mail delivery, or confirming prescription receipt at an outside pharmacy. All PEP courses were covered by patient health insurance, with co-pays ranging from $0 to $30 per medication.

Balancing Measures

Time to first ED PEP dose decreased nonsignificantly from mean time-to-dose 5.75 hours pre-intervention (range, 2.32–13.92) to 3.80 hours (range, 1.27–8.83) during intervention months (2-sided P value, .08 [data not shown]). Of 117/228 (51%) eligible physicians who completed the questionnaire, 88% rated the PEP algorithm helpful and 69% rated it easy to use.

DISCUSSION

To maximize HIV prevention in exposed patients, timely PEP is critical; delays and interruptions in medication receipt could reduce PEP’s effectiveness. To our knowledge, this study is the first to demonstrate an intervention framework to decrease pediatric HIV PEP delays. Meds-in-Hand was successfully implemented for 17 of 21 eligible individuals. Two-thirds of patients were evaluated after-hours and thus not eligible for Meds-in-Hand. However, starter pack receipt improved and PEP interruptions, particularly those caused by “preventable” insurance, pharmacy, or prescriber-related issues, decreased during the intervention. Pharmacy involvement prevented potential delays by processing prescriptions earlier and preempting insurance issues. Balancing measures revealed positive physician responses, and the intervention did not increase time to ED PEP dose. A major strength of this initiative was multidisciplinary collaboration to improve patient care.

Small sample size and only approximately 50% institutional follow-up prevented analysis of course completion or HIV transmission. However, ensuring PEP receipt is a necessary step to enable PEP completion. Data collection relied on infectious diseases consultation and pharmacy workflow activation. Additional HIV PEP cases may have been missed. Additionally, the academic medical center setting with pediatric infectious diseases personnel and a linked outpatient pharmacy may limit generalizability; however, other centers may benefit from leveraging collaborations across institutions or with outside pharmacies. Finally, our population was largely insured. A full 28-day PEP course costs up to $3000 wholesale; however, prospective clinician or pharmacist involvement could help patients apply for pharmaceutical company prescription assistance programs to reduce or eliminate costs.

CONCLUSIONS

We demonstrated that collaboration among pharmacy, infectious diseases, and pediatric emergency medicine departments improved patient care and decreased interruptions in pediatric HIV PEP. Further work should help ensure not only timely HIV PEP receipt but also completion in children and adolescents with possible HIV exposure to improve prevention efforts.

Notes

Acknowledgments. We thank the Boston Medical Center Pediatric Infectious Disease Section, Pediatric Emergency Medicine Section, Pharmacy Department, and Pediatric Department for their help in implementation and ongoing evaluation of this quality improvement initiative.

Financial support. This work was supported by the Committee of Interns and Residents, Patient Care Fund Grant, Boston Medical Center; the National Center for Advancing Translational Sciences, National Institutes of Health (NIH), through the Boston University (BU) Clinical and Translational Sciences Institute (grant1UL1TR001430); and the National Institute of Allergy and Infectious Diseases, NIH, through the BU Clinical HIV/AIDS Research Training Program (grant5T32AI052074-12). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

Potential conflicts of interest. D. F. C. receives research support from ViiV Healthcare and Merck unrelated to this study. All remaining authors: No conflicts of interest relevant to this article to disclose. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.