Home Link: Post Hospital Care to Reduce HIV Mortality in South Africa (HomeLink)

Background: This is a pilot randomized clinical trial (RCT) to demonstrate the feasibility and acceptability of a structural and behavioral intervention to reduce mortality following hospital discharge for people with HIV (PWH) in South Africa. Investigators' prior study showed that among 121 PWH discharged, 54% were readmitted and 26% had died by six months following discharge. In the prior study, investigators identified that missing clinic visits after discharge was associated with death. Here investigators are seeking to overcome key barriers in piloting a home-based post-hospital care intervention. Investigators' approach is informed by a conceptual model of key barriers to the care transition along with a behavioral explanatory model, the Behavioral Model for Vulnerable Populations. The overarching goal of this study is to tailor and pilot the intervention that shifts initial post-discharge care from the out-patient clinic to the home and provides patient-centered counseling (Home Link intervention). For the intervention to prove effective it will need to substantially reduce post-discharge mortality. Specifically, in the Home Link intervention, a team will conduct home visits to (1) provide a structured clinical assessment; (2) reconcile medications, (3) provide psychosocial support through patient-centered counseling, and (4) assess home needs (food security). These visits will start one week after discharge and be repeated every two weeks until the participant is stabilized and ready to initiate lower intensity clinic-based services or three months have elapsed. Aims: The aims of the study are to pilot a randomized clinical trial of home delivery of health services during the post-hospital period for PWH. Methods: This project is a pilot randomized clinical trial (RCT) to refine and test the feasibility, acceptability, and preliminary effectiveness of the HomeLink intervention. At the conclusion of the R34 grant period investigators will have a protocol and procedural manual ready for a full RCT powered for effectiveness. Significance: The proposed study is consistent with NIH HIV/AIDS highest priority research and the South African National Strategic Plan on HIV, tuberculosis (TB), and sexually transmitted infections (STIs) 2017-2022. The research addresses the HIV/AIDS Research Priority of "retention and engagement in these services, and achievement and maintenance of optimal prevention and treatment responses."

BACKGROUND Sub-Saharan Africa HIV mortality: In 2016 in South Africa approximately 154,000 PWH died. Of these deaths, 110,000 have been characterized as excess mortality due to HIV (HIV associated deaths). Overall, HIV-associated deaths are associated advanced disease defined by a low cluster of differentiation 4 (CD4) count. In-Hospital and Post-Hospital Mortality in Africa: Inpatient mortality among people with HIV (PWH) in Africa is uniformly high ranging from 17.2 to 29.6% across the continent, including in South Africa. Following discharge from an index hospitalization the mortality is also high. In investigators' preliminary work investigators have observed a 26% mortality in the 6 months after hospital discharge. This observation is consistent with mortality from other reports: 50% mortality at twelve months and 30% and 31% at six months reported from Tanzania, Kenya, and South Africa, respectively. Contribution of Post-Hospital Mortality to Overall Death among People with HIV: In South Africa there are approximately 84,633 public sector acute care hospital beds of which approximately 25% are adult medicine beds. By extrapolating the post-discharge mortality investigators observed to all discharges among PWH in South Africa investigators can estimate the annual mortality among PWH with a hospital discharge in the past six months. Considering the national proportion of all admissions that are for PWH, the average length of stay, and the annual readmission rate an estimated 240,000 PWH are admitted to medicine wards in South Africa annually. A six-month post-hospital mortality of 26% suggests a total of 62,374 post-hospital deaths among PWH. This number of deaths represents 40% of the approximately 154,000 total annual deaths among PWH in South Africa. A modest reduction in post-hospital deaths could substantially reduce overall HIV mortality in South Africa. Causes of Post-Hospital Death: Some post-hospital deaths may be unavoidable due to advanced or incurable disease. Other post-hospital deaths may be avoidable with improved retention in care. A study from Johannesburg, South Africa assessed loss from care for HIV-TB co-infected patients; only 42% of patients discharged were known to have been to an HIV clinic within 30 days of hospital discharge [the time needed to obtain or refill antiretroviral therapy (ART) prescriptions at the time of the study]. Similarly in Investigators' preliminary study, only 19% of participants attended a scheduled follow-up visit. Failure to attend a scheduled visit was associated with subsequent mortality in Investigators' preliminary study and in a report from Tanzania. At present, there is limited research regarding approaches to retaining PWH in care during the post-hospital transition. STUDY OBJECTIVES The overarching goal of this proposal is to demonstrate clinical trial feasibility, acceptability, and preliminary efficacy of structured post-discharge medical home visits to reduce mortality following the transition from the hospital to home. Investigators are proposing a pilot randomized clinical trial (RCT) to determine preliminary efficacy, feasibility of trial implementation and intervention delivery, acceptability of the intervention, the effect of the intervention on barriers to care, and to obtain detailed baseline psychosocial and medical needs to inform intervention refinement and explain outcomes. The primary objectives of this study are: To estimate 6 month mortality rates and assess for evidence of efficacy of the Home Link intervention compared to care as usual. To determine the feasibility, acceptability, and effect of the Home Link intervention on identified barriers to care. To characterize medical and behavioral needs among participants in the study. The secondary objectives of this study are: Determine post-hospital health care out-of-pocket costs for participants Assess re-admission Estimate 12 month mortality by study arm SIGNIFICANCE This study fits with the South African National Strategic Plan to increase effective HIV treatment. The study is also consistent with NIH HIV/AIDS highest priority research and the South African National Strategic Plan on HIV, TB, and STIs 2017-2022. Given the high post-discharge mortality findings from this study could contribute to approaches to reduce mortality for PWH. METHODS Study design: This is a pilot randomized clinical trial (RCT) to refine and test the feasibility, acceptability, and preliminary effectiveness of Home Link. Study setting: This pilot operational research study will be conducted in Matlosana. All participants will be recruited from Tshepong Hospital. Eligibility will include residing within Matlosana sub-district. Tshepong Hospital: Tshepong Hospital is a tertiary care hospital in Matlosana. An average of 30 patients are admitted to medicine wards daily. Nearly 50% are PWH. Description of the geographical areas of study implementation: The proposed study will be conducted at the 500 bed Tshepong Hospital in the Northwest Province. Tshepong Hospital is one of 5 hospitals in Kenneth Kaunda District and the single tertiary care public hospital serving the population of 400,000 in Matlosana. This hospital was selected due to its size, heterogeneous catchment area (rural to urban), and long-standing working relationships with the research team. Tshepong Hospital currently provides discharge services that include (1) discharge counseling and education provided by counselors and (2) a medical record notebook with a discharge summary for the patient to take to his or her clinic visits. Rationale for selecting geographical areas of study implementation: Investigators selected Tshepong Hospital based on a long-standing working relationship between the PI, the Perinatal HIV Research Unit (PHRU), and the Tshepong Hospital. This has included a prior study of post-hospital outcomes. Study Population: The study population is comprised of adults (≥18 years old), either male or female, spending at least two nights in a medical ward of Tshepong Hospital. Investigators will only include individuals with known HIV status, either HIV-positive or HIV-negative. Investigators propose to recruit up to 180 HIV-positive and 60 HIV-uninfected participants. Randomization to the intervention vs control arms will be stratified by HIV status. Pre-discharge study procedures: All participants will have a baseline demographic health and psychosocial questionnaire (E001, E002). Care-as-usual (CAU) study arm: Participants will receive standard discharge care as provided at Tshepong Hospital during the study. This currently includes discharge counseling from a trained discharge counselor and will be provided with a follow-up return date (usually two weeks post-hospital). Discharge counseling will include a review of discharge medications and instructions regarding follow-up care visits. Home Link study arm: The Home Link intervention will be delivered by a home visit team including a primary care nurse and counselor trained in patient-centered. A rotating hospital-based doctor will be available for pre-home visit clinical file review and post-visit discussion, via cell phone, for decision making and input on patient care during a household visit. Investigators have termed this individual a "discharge officer". The discharge officer will be a Tshepong clinician who is working in the hospital. Supporting Home Link is expected to take <30 minutes of the physician's time during the day. For study-specific concerns, the team will consult with a Good Clinical Practice (GCP)-trained, PHRU research doctor based at the Tshepong Hospital. Post-discharge follow-up: The primary outcome is mortality six months after hospital discharge. Secondary outcomes include mortality 12 months after hospital discharge, number of nights spent in a hospital, number of ambulatory clinical encounters after the index hospitalization, and patient out-of-pocket costs for medical care. Outcomes will be assessed through scheduled contact and completion of a structured questionnaire with the participant or designated next of kin (either telephonic or in person if telephonic fails) at 8, 12, 26, and 52 weeks. Mortality will be further ascertained through a 26 and 52 week review of the Tshepong Hospital file (including the affiliated Klerksdorp hospital files) and matching of participant national identification numbers to the vital statistics register through consultation with the Department of Home Affairs. Repeat hospitalization will be ascertained using self-report / next of kin report and hospital records. Most subsequent hospitalizations can be expected to be at Tshepong Hospital due to the substantial distance to reach the next nearest hospital (>50 km to a lower level facility). In Investigators' prior experience investigators have observed rare use of alternative hospitals. Clinic visit data will be ascertained by self-report followed by verification through review of clinic paper or electronic (tier.net) records and National Health Laboratory Service electronic records (TrackCare).

Participants will be randomized 1:1 to care as usual or Home Link arm. We will perform individual randomization blocked by HIV-status to enroll 30 HIV-uninfected per arm and 90 PWH per arm.

Participants and members of the research team recruiting and implementing the strategy will be unmasked to randomization assignment due to the behavioral nature of the strategy and need for team members to explain the study arm procedures to participants. Study assignments will be masked to the investigators until all outcome data have been collected.

Participants will receive standard discharge care as provided at Tshepong Hospital during the study. This currently includes discharge counseling from a trained discharge counselor and will be provided with a follow-up return date (usually two weeks post-hospital). Discharge counseling will include a review of discharge medications and instructions regarding follow-up care visits.

The Home Link intervention will be delivered by a home visit team including a primary care nurse and counselor trained in patient-centered counseling. A rotating hospital-based doctor will be available for pre-home visit clinical file review and post-visit discussion, via cell phone, for decision making and input on patient care during a household visit. We have termed this individual a "discharge officer". The discharge officer will be a Tshepong clinician who is working in the hospital. Supporting Home Link is expected to take <30 minutes of the physician's time during the day. For study-specific concerns, the team will consult with a GCP-trained, PHRU research doctor based at Tshepong Hospital.

In the Home Link intervention, a team will conduct home visits to (1) provide a structured clinical assessment; (2) reconcile medications, (3) provide psychosocial support through patient-centered counseling, and (4) assess home needs (food security). These visits will start one week after discharge and be repeated every two weeks until the participant is stabilized and ready to initiate lower intensity clinic-based services or three months have elapsed.

Acceptability questionnaire with 7 acceptability domains for healthcare workers (HCWs) and participants and in-depth interviews for HCWs and participants. An a priori code book will be developed that reflects key analytic concepts of predisposing, enabling, and need characteristics from the Andersen Behavioral Model. During the process of reading and coding of transcripts using this initial coding scheme, additional codes may be added to document emerging themes of interest.

Feasibility of clinical trial will be determined by the proportion of participants with complete follow-up.

Individual and composite score based on proportion of planned home visits completed, proportion of home medical assessments completed, proportion of home medication reconciliation completed, proportion of participants receiving patient-centered counseling, and proportion of participants with identified food insecurity who receive food supplementation/parcels at the first home visits. The proportions will be considered individually and with a composite score to be interpreted in the context of the efficacy results with a general threshold that substantially more than half of components expected to be completed need to be completed for the intervention to be considered feasible. Success will be based on a priori target of ≥90% of planned home visits completed, successful follow-up of ≥75% of participants, and vital status ascertainment of >90% of participants.

Inclusion Criteria: Tested for HIV (can be living with HIV or HIV-uninfected) Residing within Matlosana sub-district Agree to post-discharge follow-up including home visit Able to provide informed consent or, if lacking capacity at the time of recruitment (if unable to answer basic orientation questions including name, month/year, and residence location), as determined by the study team, having a next of kin able to provide informed consent Exclusion Criteria: <18 years of age Length of stay <1 night Unknown HIV status at the point of study screening and enrollment Failure by the patient or next of kin to provide informed consent to be followed up by study staff after discharge Residing outside of Matlosana sub-district Not speaking any of the languages spoken by the study team Late exclusions occurring after enrollment: Death prior to hospital discharge Hospital stay >30 days Transfer to another hospital Relocation outside of Matlosana sub-district at the point of discharge or within 7 days of discharge