Integration of PrEP services into routine antenatal and postnatal care: experiences from an implementation program in Western Kenya

Abstract

Background:

Programmatic approaches for delivering PrEP to pregnant and postpartum women in settings with high HIV burden are undefined. The PrEP Implementation for Young Women and Adolescents (PrIYA) Program developed approaches for delivering PrEP in maternal child health (MCH) clinics.

Methods:

Under the PrIYA Program, nurse-led teams worked with MCH staff at 16 public, faith-based, and private facilities in Kisumu, Kenya to determine optimal clinic flow for PrEP integration into antenatal (ANC) and postnatal (PNC) care. A program-dedicated nurse facilitated integration. HIV-uninfected women were screened for behavioral risk factors; same-day PrEP was provided to interested and medically-eligible women. PrEP and MCH services were evaluated using standardized flow mapping and time-and-motion surveys.

Results:

Clinics developed two approaches for integrating PrEP delivery within ANC/PNC: 1) co-delivery: ANC/PNC and PrEP services delivered by same MCH nurse or 2) sequential services: PrEP services after ANC/PNC by a PrEP-specialized nurse. Three clinics selected co-delivery and 13 sequential services, based on patient volume and space availability. Overall, 86 ANC/PNC visits were observed. Clients who initiated PrEP took a median of 18 minutes (IQR 15–26) for PrEP-related activities (risk assessment, PrEP counseling, creatinine testing, dispensation, and documentation) in addition to other routine ANC/PNC activities. For clients who declined PrEP, an additional 13 minutes (IQR 7–15) was spent on PrEP-related risk assessment and counseling.

Conclusions:

PrEP delivery within MCH used co-delivery or sequential approaches. The moderate additional time burden for PrEP initiation in MCH would likely decline with community awareness and innovations such as group/peer counseling or expedited dispensing.

INTRODUCTION

To protect women and reach global targets for elimination of mother-to-child transmission of HIV, it is critical to integrate HIV prevention into antenatal (ANC) and postnatal (PNC) care.^1–4 The World Health Organization (WHO) recommends that in high HIV burden settings, and for sub-populations with high (>3%) HIV incidence in low burden settings, HIV-negative women should be offered prevention interventions, including pre-exposure prophylaxis (PrEP), at ANC/PNC visits.^5–7 Programmatic delivery of PrEP for pregnant and postpartum women is being considered in high-prevalence regions⁸; however, implementation approaches have not been defined. Similar to operational research that optimized prevention of mother-to-child HIV transmission (PMTCT) programs ^9,10, research is needed to optimize PrEP delivery to this unique population.⁴

Routine MCH clinics are attractive settings for PrEP implementation.⁴ Attendance at ANC/PNC is high in sub-Saharan Africa¹¹, women are universally offered HIV testing during ANC¹², and PMTCT services are integrated into ANC/PNC.¹³ Leveraging existing HIV testing and ART infrastructure within MCH clinics for PrEP delivery holds promise as a feasible delivery approach with high coverage and efficiency.

PrEP Implementation for Young Women and Adolescents (PrIYA) is an implementation program that is currently delivering PrEP to women seeking routine ANC/PNC services at 16 facilities in Kisumu County, Kenya.¹⁴ In the current operational investigation, we defined approaches for integrating PrEP into routine ANC/PNC using the PrIYA Program as a case study in order to inform best practices for operationalizing PrEP delivery within the context of ANC/PNC. The objective of this paper is to evaluate the work flow patterns and additional staff time associated with integrating PrEP into ANC/PNC services.

METHODS

PrIYA Program and setting

The PrIYA Program is a two-year implementation project to evaluate PrEP delivery strategies to reach adolescents and young women at high risk for HIV acquisition in the context of public sector maternal and child health (MCH) and family planning systems.¹⁴ PrIYA is part of the DREAMS Innovation Challenge funded by PEPFAR managed by JSI Research & Training Institute, Inc. In collaboration with Department of Health and Sanitation, Kisumu County and the National AIDS and STI Control Programme (NASCOP), PrIYA has been implemented in 16 facilities (11 public, 4 faith-based, 1 private) in Kisumu County, Kenya, a region with adult HIV prevalence of 19.9% (up to 28% among pregnant women).^15–17 The current evaluation focuses on PrEP initiation within the context of ANC/PNC clinical service delivery.

Service delivery organization and clinic flow mapping

Within the PrIYA Program, 40 program-supported nurses were trained on clinical PrEP delivery per national guidelines, assigned a work station at a public or private sector MCH clinic and advised to integrate PrEP within the ANC/PNC clinic flow. Program nurses initially worked with clinic staff, county-level health officials and the Kenya Medical Supplies Agency, to ensure PrEP commodities were in place. Starting in June 2017, PrIYA program-dedicated nurses began providing PrEP. Nurses screened HIV-uninfected pregnant and postpartum women for willingness to consider PrEP and behavioral risk factors to inform PrEP counseling per NASCOP guidelines.⁸ Medically eligible women who wanted to initiate PrEP received same-day PrEP. All PrEP-specific services were delivered by nurses within the ANC/PNC clinic, including medication dispensation, at no additional cost to clients.

To define the sequence of activities in each facility’s integrated PrEP delivery, we engaged nurses delivering PrEP services at each facility. Each nurse drew maps, without a pre-determined structure, of the flow of clients through their facility from the client’s arrival at the facility to departure. Maps were drawn of the clinic flows with and without PrEP services. Activities, both clinical (counseling, physical examination, laboratory testing) and non-clinical (registration, waiting, walking to examination rooms) from one step to the next were mapped. Based on flow maps, activities were grouped into categories reflective of the core components of ANC/PNC and PrEP services (Table 1).

Table 1.

Activities and time spent during clinic visit with integrated ANC/PNC and PrEP services

Category	Activity	N	Time in minutes
			Mean (Min-Max)	Median (IQR)
Pre-service	Waiting to receive services, registration	44	22 (3–75)	13 (7–36)
HTS	HIV testing and counseling	15	13 (4–24)	12 (10–16)
ANC/PNC1	ANC services	11	16 (4–52)	9 (8–13)
	Lab (no HTS or creatinine)	9	38 (20–60)	40 (30–43)
	PNC services	25	9 (2–28)	6 (4–13)
PrEP-specific	Behavioral risk assessment and PrEP counseling2
	Declined PrEP	10	10 (2–15)	13 (7–15)
	Willing to initiate3	14	19 (6–45)	15 (14–23)
	POC creatinine testing4	16	4 (1–8)	3 (2–4)

¹ANC/PNC services include physical examination, well-child visits, childhood vaccination, laboratory testing (syphilis and hemoglobin)

²All clients received assessment for behavioral risk (completion of a questionnaire on behavioral risk factors for HIV) and willingness to consider PrEP. Initial assessment also included general informational counseling on PrEP depending on the client’s awareness and interest in PrEP.

³Among clients who were willing to initiate PrEP, PrEP counseling also included information on how to use PrEP and adherence as well as medication dispensation

⁴Only among clients willing to initiate PrEP, creatinine testing using Nova StatSensor® Xpress(TM) Creatinine point-of-care handheld analyzer ³⁸ was conducted.

In addition to clinic flow mapping, we collected information on nurse staffing and client volume at each clinic to provide insight on factors influencing each clinic’s PrEP integration approach. Nurses delivering PrEP were asked to describe the primary reason the integration approach was adopted within their respective ANC/PNC clinics.

Time-and-motion data collection and analysis

To quantify time required to deliver PrEP services, we enumerated the visit times for ANC and PNC clients at 6 facilities. At the time of enumeration, nurses had delivered PrEP to clients daily for the previous 6 months. Facilities were selected based on their proximity to Kisumu city (<30 minutes driving time.) Trained observers timed ANC/PNC visits using a stopwatch and a semi-structured recording sheet, allocating time to the categories in Table 1. Observers recorded times at each service component until the heterogeneity between visit durations was captured (at least 9–10 encounters). A total of 40 person-hours of observations were recorded. During observations, the observer talked with nurses when necessary, but did not provide any feedback and did not interact with ANC/PNC clients. We calculated summary statistics (mean, median, and interquartile range [IQR]) for time spent in each category and estimated the additional PrEP-related service time.

RESULTS

Approaches of integrating PrEP delivery within ANC/PNC

Clients’ key activities during routine visits were distilled from clinic flow maps. In most facilities, laboratory personnel or counsellors conducted HIV testing services (HTS) directly after registration and a nurse subsequently provided ANC/PNC services with the client’s HIV status known. However, in some facilities, clients received HTS and ANC/PNC services from a single nurse. With integration of PrEP into ANC/PNC, two clinic flow approaches emerged (Figure 1a): 1) ANC/PNC and PrEP services delivered by the same nurse at the same time (co-delivery approach), or 2) ANC/PNC services delivered by an MCH nurse followed by PrEP services delivered by a nurse with PrEP expertise in a separate location (still within the ANC/PNC clinic) (sequential-delivery approach). Under a co-delivery approach, PrEP-specific activities were added to the ANC/PNC client visit by the same nurse who provided ANC/PNC services and, at times, also performed HTS. Under a sequential-delivery approach, a client first received HTS prior or during ANC/PNC services from an MCH nurse. If HIV-negative, the client was referred to a second nurse with PrEP-specific expertise for PrEP-related activities. PrEP-specific activities in this approach were also delivered within the ANC/PNC clinic, though typically in a separate space, allowing other clients to continue with ANC/PNC services. In both approaches, PrEP drugs were dispensed directly by nurses who provided PrEP-related counseling and did not require an additional pharmacy visit.

Figure 1a. Organization of activities during ANC/PNC with integrated PrEP services, under co-delivery or sequential-delivery approaches.

HTS=HIV testing services; ANC/PNC=antenatal care/postnatal, care

¹Among clients who were willing to initiate PrEP, PrEP counseling also included information on how to use PrEP and adherence as well as medication dispensation

Characteristics of facilities with integrated PrEP delivery within ANC/PNC clinics

PrEP delivery approaches were not assigned a priori, but were developed organically based on clinic volume, space, and staffing with input from leadership at each respective facility. For example, some facilities had limited clinical space for ANC/PNC clinical services but had an available space (i.e. re-purposed closet or other non-clinical room) for PrEP services. In these clinics, it was more practical to adopt a sequential-delivery approach since PrEP services could occur elsewhere within the clinic while regular ANC/PNC services continued. In other facilities, space was limited, or the client-to-provider ratio demanded all nurses deliver ANC/PNC services. Therefore, a co-delivery approach was adopted. ANC/PNC clinics typically adopted one approach, though approaches were not static and at times were modified based on circumstances within clinics. Figure 1a presents the facility, PrEP delivery approach and the primary rationale for approach.

Time spent on PrEP delivery within ANC/PNC visits

We observed 86 clients during their ANC/PNC visits. Clients spent a median time of 13 minutes (IQR 7–36, n=44) waiting to be seen by ANC/PNC nurses with high variability. ANC services lasted a median of 9 minutes (IQR 8–13, n=11), with an additional 40 minutes (30–43, n=9) spent receiving laboratory services, if indicated. Rapid plasma reagin [RPR] testing for syphilis and hemoglobin testing are indicated for clients attending first ANC visits per Kenyan guidelines. PNC visits were shorter, lasting a median of 6 minutes (IQR 4–13, n=25)). Clients spent a median of 12 minutes (IQR 10–16) receiving HTS (n=15).

Time spent on PrEP-specific activities depended on whether the client initiated PrEP. PrEP education and counseling lasted a median of 13 minutes (IQR 7–15) for clients who eventually declined PrEP (n=10). Overall, clients willing to initiate PrEP (n=14) spent a median of 18 minutes (IQR 15–26) on PrEP-specific activities, including counseling, focused coaching on how to use PrEP and adherence, medication dispensation, point-of-care (POC) creatinine testing and documentation. POC creatinine testing alone took a median of 3 minutes (IQR 2–4, n=16) and PrEP-specific activities excluding POC creatinine testing lasted a median of 15 minutes (IQR 14–23). All clients observed who desired PrEP were medically eligible and received PrEP directly from nurses at the same visit.

DISCUSSION

Within this first large-scale PrEP implementation project for pregnant and postpartum women in a high HIV burden setting, clinics developed two general approaches for integrating PrEP delivery within ANC/PNC. PrEP-specific activities took <20 minutes per client, which could translate to several additional hours of work for MCH nurses per week, depending on client volume. However, time required for one-on-one informational counseling would be expected to decrease as community-level awareness of PrEP increases and as new approaches for increasing efficiency of PrEP delivery are introduced.

We found that initiating clients on PrEP within ANC/PNC requires a moderate amount of additional time per client within already heavily burdened MCH systems. ^18–21 We did not use group or peer counseling approaches, which could reduce time spent by healthcare workers (HCWs). Group counseling is an effective model for delivering ANC-related information and has been shown to improve overall quality of ANC. ^22–27 There is some evidence that including HIV and STI prevention messaging within group ANC counseling is associated with reduced incidence of STIs and repeat pregnancies. ^28,29 Leveraging group counseling activities within ANC/PNC to provide information on PrEP prior to one-on-one clinical services could help streamline PrEP integration. Group counseling in ANC could provide information about factors that may influence PrEP uptake specific to pregnancy and postpartum, including varied levels of sexual activity ^30–33, managing PrEP side effects, and addressing concerns about the drug effects on infants. ³⁴

In sub-Saharan African countries, HIV testing is routinized during ANC ¹² and repeat HIV testing during ANC/PNC is recommended. ³⁵ However, clinics newly integrating PrEP into ANC/PNC could initially experience increased volume of HTS clients, which may cause service bottlenecks. HIV self-test kits could be used by clients while waiting to receive ANC/PNC services or used by clients at home to reduce clinic-based HTS burden. To date, studies exploring the utility of HIV self-tests within the context of ANC/PNC focus solely on promoting male partner HIV testing.^36,37

In Kenya, PrEP initiation guidelines recommend, but do not require, testing for creatinine clearance (CrCl) at PrEP initiation and annually thereafter if the clinic has laboratory capacity to do so.⁸ Under the PrIYA Program, prior to PrEP initiation, nurses conduct creatinine serum testing using validated Xpress StatSensor® POC machines ³⁸ and calculate CrCl using a mobile application. POC creatinine testing added <5 minutes to visit time. However, in settings where CrCl is required by national guidelines or where POC testing is unavailable, routine monitoring may require significantly more time. Many PrEP programs have a high rate of early discontinuation among individuals who initiate PrEP.³⁹ Thus, Cr testing in the subset who continue PrEP use for >1 month (or 6–12 months) may be one way to streamline this process.

Our assessment has limitations. Two integration approaches emerged from facilities, however, there may be other approaches that facilities develop in the future. Our current evaluation was not designed to quantitatively compare integrated delivery approaches and our statistical power to do so was very limited. Optimizing operational success of PrEP delivery within MCH settings will require additional evaluations that formally test differences between approaches and assess acceptability among staff and patients. Our evaluation focuses on PrEP initiation. Future studies are needed to understand implications of PrEP follow-up visits on clinic workload within the context of ANC/PNC. We included 86 observations in our time-and-motion evaluation, though some activities had <20 observations. Time-and-motion studies are labor-intensive and are therefore commonly done with small samples. ^40,41 Time spent waiting and for specific services varied by facility. The absolute time devoted to PrEP activities may decrease with programmatic innovation or experience. The first phase of implementation is often time-intensive and a number of innovations for counseling, PrEP screening, and PrEP dispensation and documentation could decrease the time requirement. With wider awareness of PrEP, behavioral screening and counseling time would decrease and become more automated. Women who successfully navigate PrEP use could potentially serve as peer counselors, similar to peer counselors in PMTCT, and provide more relevant adherence counseling at a lower wage cost and leaving existing MCH HCW available for routine ANC/PNC services.

CONCLUSIONS

We observed two successful clinic approaches for integrating PrEP delivery within routine ANC/PNC with moderate additional time required for initiating willing clients on PrEP. Innovative approaches for increasing efficiencies and PrEP follow-up could further optimize PrEP delivery in ANC and PNC in high HIV burden settings.

Figure 1b. PrEP integration approaches and characteristics of 16 facilities included in the PrIYA Program.

¹ Within the Kenyan health system, public sector facilities are classified from levels 1–5 which corresponds to overall client volume and capacity to deliver tertiary care. Level 1 corresponding to the lowest volume primary care facilities whereas Level 5 corresponds to large volume referral hospitals with multiple clinical services available.

² The average number of HIV-uninfected ANC/PNC clients per week attending each clinic was calculated based on aggregated tallies reported daily as part of the PrIYA Program

³ The number of ANC/PNC nurses with PrEP expertise are defined as nurses who have received specialized training and mentorship under the PrIYA program.