Training Mid-Level Providers to Treat Severe Non-Communicable Diseases in Neno, Malawi through PEN-Plus Strategies

Background: Non-communicable diseases (NCDs) are a leading cause of worldwide morbidity and mortality, yet access to care in lower-income countries is limited. Rural communities, where poverty levels are high, feel the greatest burden. In Malawi, as elsewhere in the African region, it is particularly challenging for patients in rural districts to obtain care for locally endemic and severe NCDs such as type 1 diabetes, rheumatic heart disease, and sickle cell disease. The Package of Essential NCD Interventions – Plus (PEN-Plus) is a strategy to decentralize care for these severe conditions by enabling local clinicians at intermediate-care facilities to provide services otherwise available only through specialty clinics at central hospitals. Objectives: The primary objective of this study was to evaluate the impact of training mid-level providers to treat severe and chronic NCDs in newly established PEN-Plus clinics in Neno, Malawi. Methods: Our team developed a logic model to describe the anticipated impacts of the intervention on provider knowledge, patient recruitment, and care provision. We applied a retrospective review of routinely collected clinical and administrative data to assess changes along these hypothesized pathways. Findings: Didactic trainings improved provider test scores immediately following training (25-point improvement; p < 0.01), with demonstrated retention of knowledge after 6 months (21-point improvement, p < 0.01). Over 350 patients were enrolled in the first 18 months of program initiation. The PEN-Plus clinic led to significant improvement in the provision of medications and testing across a range of services. Conclusion: Mid-level providers can be successfully trained to treat severe NCDs with physician-guided education, mentorship, and supervision. The PEN-Plus clinic improved patient enrollment, the quality of clinical care and access to essential medications and laboratory supplies. These lessons learned can guide decentralization of NCD care to district hospitals in Malawi and expansion of PEN-Plus services in the African region.

This study is a retrospective review of routinely collected clinical and administrative data. Using the Implementation Research Logic Model [16], our team developed a logic model to structure our review (Figure 1). Using available data, we focus on three mechanisms (provider knowledge, patient recruitment, and care provision). Theory of change, highlighting three mechanisms of action. Neno is a rural district of 138 291 people, located in the southern region of Malawi [17]. The majority of people rely on subsistence farming and less than five percent of the population has electricity [18]. Poor transportation infrastructure and road maintenance makes it difficult for patients to access hospitals and patients typically travel long distances on foot to reach health care centers. Neno’s local health system is mostly public and is comprised of a district hospital in the mountainous western region, a community hospital in the lower, eastern region and 12 health care centers. It is often difficult for patients in rural districts in Malawi, such as Neno, to obtain adequate care for complex NCDs. Specialized clinics for insulin-dependent diabetes, sickle cell disease and chronic cardiac failure only exist at the 4 central hospitals. Travel to a central hospital from Neno takes hours and is unaffordable for most patients. At the district level, midlevel providers, including clinical officers and nurses, provide most of the treatment at chronic care clinics. However, they are widely not trained or given resources to provide care for complex NCDs and minimal physician support is available. Partners In Health (PIH), a non-government organization, has partnered with the Ministry of Health (MOH) in Neno since 2007. PIH combines accompaniment and direct service provision with a goal of overall health systems building. Major areas of support include HIV care, maternal health, community health and non-communicable diseases. Since 2009, the MOH and PIH have worked to strengthen care for NCDs in the district, first with the introduction of Chronic Care Clinics (CCC) treating patients diagnosed with a variety of NCDs at the 2 hospitals, and then through the Integrated Chronic Care Clinics (IC3). IC3 clinics were decentralized and led by mid-level providers and provide outpatient services for HIV and non-communicable chronic conditions at weekly clinics located at health centers across the district [19,20]. As part of this effort, the team has worked to address gaps in staffing, equipment and medicines. Partners In Health, in conjunction with the MOH, planned on opening two PEN-Plus clinics in October 2018. Weekly clinics were scheduled at Neno District Hospital and Lisungwi Community Hospital. These clinics would provide care for severe NCDs while the Integrated Chronic Care Clinics (IC3) would maintain services for basic NCDs included in the PEN package such as hypertension, type 2 diabetes, chronic respiratory disease, epilepsy, basic mental health conditions, and cervical cancer screening. In anticipation of opening the clinics, PIH held a series of NCD trainings for clinicians and nurses in Neno District from August 2018 to February 2019. All PIH and MOH clinical officers and nurses in the district were invited to attend the trainings to improve case finding and district-wide treatment of NCDs. However, an essential purpose of the trainings was to prepare the midlevel providers who would work in the PEN-Plus clinics. A total of 4 sessions over 7 days were given in diabetes, cardiovascular disease, pulmonary and renal disease, and gastrointestinal, neurologic and hematologic disease. Trainings were a mix of clinical lectures, case studies and strategies on counseling patients. Materials were developed by internal medicine, cardiology, and pulmonary specialists at an appropriate level for mid-level providers. Didactic PEN-Plus trainings were supplemented with in-clinic mentorship. At the PEN-Plus clinic, a physician served as a clinical mentor for the NCD clinical officers, and a nurse mentor supervised the NCD nurses in patient care and counseling. The mentors would see patients with staff in the clinic which provided the opportunity for onsite, real-time questions and feedback. Physician led mentorship started weekly, but gradually decreased to bi-weekly and monthly as the clinical officers became more proficient over the first year of the clinic. The essential staff hired for each clinic include a clinical officer, nurse, clerk, and pharmacy assistant. Stock of some vital NCD medications, such as insulin, ACE-inhibitors, and beta-blockers, were supplemented in IC3 prior to establishment of PEN-Plus clinics, meaning that changes in medication distribution are not attributed to increased availability. Clinical knowledge was assessed using structured 40-point questionnaires given at 3 points: immediately prior to training, immediately following the initial training, and 6 months post training. Tests were separately developed for each daily NCD session and included an assessment of epidemiology, diagnosis, management, medication, and counseling. We conducted paired t-tests to assess for differences in scores across the 3 time points. To investigate the care given at the PEN-Plus clinic since inception, we examined process outcomes for selected advanced NCDs. Using electronic medical records, we gathered data on the proportion of patients receiving medical tests and appropriate medications between January 2017 and January 2020 – a period covering 18 months before and after the clinic was established. For all eligible visits, we collected patient-level (age, gender, HIV-comorbidity) and visit-level (lab tests, medication, symptoms) data from the clinic’s routine data system. For the purposes of this work, we focused on insulin-dependent diabetes, chronic kidney disease and chronic heart failure as they are among the most common conditions seen at the PEN-Plus clinics. Data collected prior to the PEN-Plus clinic was collected from the IC3 clinic in Neno with both clinics using the same protocols for the outcomes considered. Hence, any increase in the distribution of testing or medication use is expected to be attributed to the training, mentorship, dedicated staff, and clinic for complex NCD patients, rather than systematic differences in the patient population. Chi-square tests, univariate, and multivariable logistic regression were used to assess for differences in the provision of tracer drugs between the groups. Multivariable models control for patient-level characteristics (age, gender, and HIV-comorbidity) that are hypothesized to affect the care [21]. To ease interpretability of our results, we use the “margins” command to translate these results into predicted probabilities of correct clinical care with and without training. All analyses were conducted using Stata version 15 [22]. Diabetes and chronic kidney disease analysis focused on lab testing per recommendations. Clinical guidelines for patients with insulin-dependent diabetes recommend blood sugar testing during each clinic visit and Hemoglobin A1C (HbA1C) testing once per 6 months. Patients with chronic kidney disease are recommended to receive a urine protein level every 3 months and a blood creatinine level every 6 months. Chronic heart failure review looked at appropriate medication use. Patients with a diagnosis of New York Heart Association (NYHA) class II or greater are recommended to be on an ACE Inhibitor and Beta Blocker. Additionally, patients with a diagnosis of chronic heart failure and symptoms of shortness of breath, lower extremity edema or fluid overload are recommended to be on furosemide. This study was reviewed and approved by the Neno District Health Research Committee. The study falls under the PIH umbrella IRB with the Malawi National Health Sciences Research Committee (NHSRC) as regularly collected patient data was used.

• ruderman, todd
• chibwe, evelyn
• boudreaux, chantelle
• ndarama, enoch
• wroe, emily b.
• connolly, emilia
• bukhman, gene