Background Neonatal sepsis is a leading cause of mortality, yet the recommended inpatient treatment options are inaccessible to most families in low-income settings. In 2015, the World Health Organization released a guideline for outpatient treatment of young infants (0-59 days of age) with possible serious bacterial infection (PSBI) with simplified antibiotic regimens when referral was not feasible. If implemented widely, this guideline could prevent many deaths. Our implementation research evaluated the feasibility and acceptability of implementing the WHO guideline through the existing health system in Malawi. Methods A prospective cohort study was conducted in 12 first-level health facilities in Ntcheu district. Trained health workers identified and treated young infants with PSBI signs with injection gentamicin for 2 days and oral amoxicillin for 7 days, whereas those with only fast breathing were treated with oral amoxicillin for 7 days. Health Surveillance Assistants (HSAs) were trained to promote care-seeking and to conduct home visits on day 3 and 6 to assess infants under treatment, encourage treatment adherence and remind the caregiver to return for facility follow up. Infants receiving outpatient treatment were followed up at health facility on day 4 and 8. The primary outcome was proportion of outpatient cases completing treatment per protocol. Findings A total of 358 infants received outpatient treatment (202 clinical severe infection, 156 only fast breathing) from February to September 2017. Of these, 92.7% (332/358) met criteria for treatment completion and 88.8% (318/358) completed the day 4 follow-up. Twelve (3.4%) young infants clinically failed treatment with no reported deaths in those treated at outpatient level. This treatment failure rate was lower than those reported for the simplified regimens tested in the SATT (8-10%) and AFRINEST (5-8%) equivalency trials. More than half of infants (58.1%; 208/358) received HSA follow-up visits on days 3 and 6. Conclusion Study results demonstrate the feasibility of outpatient treatment for sick young infants when referral is not feasible in Malawi, which will inform scale-up in other parts of Malawi and countries with similar health system constraints.
This implementation research was prospectively conducted in a cohort of sick young infants. The primary objective was to evaluate the feasibility of treatment of PSBI for young infants up to 2 months of age delivered through first-level care facilities in Malawi when referral was not feasible. Secondary objectives were to: i) evaluate whether first-level health facilities can deliver quality care for infants with PSBI and fast breathing only, ii) assess the extent to which Health Surveillance Assistants (HSAs) can follow-up young infants with signs of PSBI or fast breathing only in the community, and iii) assess the acceptability by families of such treatment offered at first-level health facilities. We followed the StaRI standards for reporting implementation research studies [8]. The study was conducted from February to September 2017 in first-level health facilities in Ntcheu district, located in central Malawi along the border with Mozambique. In 2016 Ntcheu had an estimated population of 0.59 million [9]. Government health services in the district are provided through three main levels: 1) referral level through the Ntcheu district hospital; 2) outpatient services through 39 first level health centres, including 28 with maternity services; and 3) village level through community health workers called HSAs. At the time of the study, young infants presenting to first-level health facilities with signs of PSBI in Malawi were referred to the district hospitals for inpatient treatment with a combination of injectable gentamicin and penicillin/ampicillin for up to 7–10 days, in line with WHO recommendations [10]. The district hospitals faced challenges to manage sick young infants for the whole district. In addition, families incurred significant costs associated with inpatient treatment such as lodging, purchasing of food, transport and other costs. The Malawian Ministry of Health (MoH) was concerned about the limited access to treatment for sick young infants and, informed by the publication of the WHO guideline for management of PSBI among young infants when referral is not feasible, expressed interest in introducing the new guideline to Save the Children (SC). The MoH, SC US and WHO partnered to introduce the WHO PSBI guideline in Malawi. Several meetings were held at national level to review the guideline and discuss the process of gradual and structured introduction. It was agreed that before introduction of the guideline across the country, implementation research would be conducted to evaluate the feasibility of case management at first-level health facilities when referral is not feasible. Ntcheu was selected by the MoH as the study district. The main reason for selection of Ntcheu was that the Community-Based Maternal Newborn Care (CBMNC) package, updated in 2014 to reflect the WHO Caring for Newborns at Home recommendations, was being introduced in this district and the MoH believed that the pregnancy and postnatal visits by HSAs under the revised CMBNC package would complement the implementation research [11–12]. A timeline of the study development and implementation milestones is provided in S1 Table. The MoH along with technical experts from WHO and SC met with the Ntcheu District Executive Committee and the District Health Management Team (DHMT) to brief them on the WHO guideline and to gauge their interest in conducting the implementation research. The team visited the community and the health facilities to assess current treatment approaches and care seeking behaviour for sick young infants. Following this exercise, the MoH, with support from SC and WHO, developed an implementation research protocol for the “Treatment of Young Infant Infections in Ntcheu” (TYIIN) study. The MoH IMCI unit led the process of adapting the protocol, training materials and data collection tools for Malawi and engaging with district stakeholders. A district-based team was formed to complete these tasks and to support the implementation of the study, including training of health workers and subsequent monitoring and quality assurance. The team comprised 13 members including IMCI master trainers, the district hospital research coordinator, programme coordinators of CBMNC, Safe motherhood, IMCI, and nurses and clinicians working in the maternity, children’s ward and the under-five outpatient department (OPD). Save the Children and the MoH also engaged community structures to raise awareness of the study and seek their inputs on strategies to mobilize communities. A stakeholder meeting was held in Ntcheu District in April 2016 to inform the chiefs from the Traditional Authorities and other non-governmental organizations working in Ntcheu on the study objectives and methods and seek their views on the acceptability of the study by the community members. Community leaders raised concerns regarding the capacity of health facility staff and HSAs to manage the sick young infants and whether families would come back for follow-up and adhere to treatment. In response, participants were assured that comprehensive training would be offered to all health facility staff and HSAs involved in the assessment and management of the sick young infants and that the study team would closely monitor all children receiving treatment. At the end of the meeting, the study was endorsed by all community leaders. Twelve outpatient facilities and their associated HSAs were selected in October 2015 in collaboration with the district health office (DHO) based on ‘best fit’ with the following criteria: presence of providers trained in IMCI; catchment population size (aiming for >15,000); proximity to the district centre to facilitate ongoing follow-up and supervision; and mix of government and Christian Health Association of Malawi (CHAM) facilities to reflect the distribution of health facilities in Ntcheu and other districts of Malawi. The sample of 12 facilities included nine government and three CHAM facilities. Facilities managed by CHAM, a large non-governmental provider of health services, are coordinated by the MoH and charge user fees to cover costs of consultation and medical supplies. The MoH has a service level agreement with CHAM to provide maternal and newborn care services free of cost for up to six weeks post-partum, which was extended to up to two months for the study. Government health services are provided free of cost. Altogether the 12 study facilities served a population of approximately 0.26 million, with an estimated 9,700 annual live births and an estimated 970 cases of PSBI annually (based on 10% incidence) [13–16]. Each facility was staffed with at least one medical assistant/clinical officer and two nurse-midwives and managed community-based services by 10 to 15 HSAs (total of 147 HSAs across study facilities). Sick young infants with clinical signs of PSBI either self-referred by the families or identified in the community by HSAs presented to the study first level health facilities. All infants with PSBI signs presenting to study health facilities were referred to the district hospital. If referral was not feasible, they were classified as only fast breathing in 12 hours (Renal failure); Cellulitis or abscess at injection site (only in those receiving injections) Trained facility staff provided outpatient treatment to sick young infants at study health facilities according to the WHO guideline (Table 1). Health facility staff administered the first dose of gentamicin injection and the first dose of dispersible oral amoxicillin on first day of treatment and the second dose of gentamicin injection the following day. Caregivers (usually mothers) were provided the remaining doses of dispersible oral amoxicillin and counselled to give to the infant twice daily until the end of the 7-day treatment period. Health facility staff showed caregivers how to prepare dispersible oral amoxicillin and taught them about what danger signs to look for and what actions to take if they noticed any danger signs. Outpatient cases were followed up on day 4 and day 8 at the facility and on day 3 and 6 in the community by trained HSAs to reinforce treatment adherence and follow-up. The health facility staff informed the HSAs in their catchment area about PSBI cases within 24 hours through phone calls, short message service (SMS), in-person or by giving the caregiver a form to hand over to the HSA. Caregivers were provided with a follow-up and medicine reminder card on the first day of treatment tailored to case type (CSI or only fast breathing) to encourage adherence to treatment and follow up. Infants who did not return for follow-up on day 4 or day 8 as advised were followed up by TYIIN study staff. Infants with critical illness or who accepted referral to hospital were followed up by TYIIN study staff for outcome assessment at day 14. Two IMCI master trainers from Malawi received training from WHO facilitators on an updated young infant IMCI module based on the 2015 WHO PSBI guideline [3]. They trained the Ntcheu district facilitators who in turn trained the staff from the 12 study health facilities at the district hospital. Nurses and clinicians responsible for managing sick young infants at the 12 study facilities, who had been previously trained in IMCI, were trained the WHO PSBI guideline, study standard operating procedures (SOPs), data collection tools, consent process and documentation, and caregiver reminder cards. Facility staff were also provided with and trained on the use of pulse oximeters to assess oxygen saturation. Clinical standardization assessment exercises were included in the training. HSAs and the health facility staff received joint training in CBMNC adopted for Malawi from the draft WHO Caring for the Newborn at Home training manual to promote coordination and team work during implementation [11–12]. This training aimed to refresh the HSAs on their role in counselling families on newborn danger signs during pregnancy and postnatal home visits and assessment and referral of sick young infants through CBMNC. The training also oriented the HSAs to the study SOPs and their additional study specific roles to follow-up on treatment adherence and completion on day 3 and 6 for enrolled cases. Facility staff received a refresher training on clinical assessment, including use of pulse oximetry for assessing oxygen saturation. The study also oriented existing 321 community volunteers, referred to as secret mothers ‘amayi achinsinsi’ who were identified by the community as an essential group of women to link the HSAs with the community. Secret mothers are elderly women selected by their communities to serve as a link between women and the health care system, particularly during pregnancy and the postnatal period [18]. The secret mother approach was introduced in 2012 as part of the Safe Motherhood Initiative and their role in linking pregnant women and mothers to HSAs and health facilities had already been quite well established in Ntcheu district at the time of our study [18]. Secret mothers were oriented to encourage mothers to seek routine postnatal services at the health facility, to raise awareness of newborn danger signs and appropriate care seeking and on their role of supporting HSAs in reminding the caregivers whose infants were receiving outpatient treatment to complete their treatment and return to the health facility for follow up. The study was implemented for seven months, following an initial two-month pilot phase. Implementation of the study was led by MoH with evaluation and measurement support from the TYIIN study team (see S2 Table for details on study roles and responsibilities). The TYIIN study team comprised four full-time staff: a manager responsible for overall coordination, two clinical monitors and a data manager. Technical assistance was provided by WHO, SC and the Malawi national IMCI Unit. In addition, the district-based team including members of the Ntcheu DHMT and supervisors from Ntcheu District Hospital supported implementation through trainings, quality monitoring, supportive supervision and mentorship plus any other study-related activities from start to completion. Prior to initiation of the study, all study facilities were assessed by TYIIN study staff using a facility preparedness checklist to determine availability of essential medicines and equipment and any missing supplies were replenished. During implementation, the TYIIN clinical monitors visited each facility monthly and replenished the supplies, as needed. In addition, facilities were encouraged to notify study staff any time they were approaching minimum stock levels. In case of stock-outs, SC replenished essential study medicines including injection gentamicin, dispersible oral amoxicillin, and injection ampillicin, Monetary incentives were provided to health facility staff and HSAs involved in the study to compensate for the additional data collection associated with the study. A total of 7,500 kwacha (about $10) per verified outpatient case was given to the health facility in charge who was responsible to share it among all clinical staff and HSA. In addition, the study also provided 5,000 kwacha (about $5) of airtime per facility per month to facilitate communication with study staff and maintain link with HSAs for follow-up. Study monitoring was conducted by three MoH district staff and the TYIIN study team. Each study facility was assigned to a MoH clinical supervisor for the purposes of study monitoring. These MoH clinical supervisors were responsible to remain in weekly contact with their assigned health facility staff through phone calls and visits where possible. A WhatsApp group was set up for MoH clinical supervisors to update the study team on the enrolment progress of their facilities and discuss study-related issues. The TYIIN clinical monitors and study manager independently supervised study facilities on a monthly basis to verify enrolment, treatment, and assess outcomes. In addition, study review meetings attended by the IMCI Unit, the DHMT, the district study team, health facility nurses and clinicians and HSAs supervisors were held monthly to track progress and address any challenges. The primary outcome was the proportion of young infants with CSI or only fast breathing aged 0–59 days completing outpatient treatment per protocol. Secondary outcomes included proportion of eligible cases accepting outpatient treatment (refusing referral), completion of day 4 follow up, completion of HSA follow up visits, treatment failure and outcome of the illness at day 14 (Table 1). Health providers checked the adherence to treatment by observing the amoxicillin blister packs on day 4 and 8 facility follow-up visits and by interviewing the child’s primary caregiver. The HSAs also tracked treatment adherence during their scheduled contacts with the family on day 3 and day 6 using a similar case recording form. Health provider records were used to determine treatment adherence for study purposes. The TYIIN study clinical monitors and study manager independently validated treatment completion for at least 15% of cases. For cases that failed to return to the health facility for the mandatory day 4 and day 8 follow-up visits, TYIIN clinical monitors and the study manager directly visited the families at their homes to determine treatment adherence and completion. Outcome assessments on day 14 were completed by TYIIN study personnel through home visits or phone calls to families. We also collected qualitative data on feasibility and acceptability through in-depth semi-structured interviews with purposively selected DHO staff, health facility staff, HSAs and mothers of infants receiving outpatient treatment. Interview guides for DHO, health facility staff and HSAs captured experiences and perspectives on successes and challenges in the identification, treatment and follow-up of sick young infants in first level facilities and the community, completing follow-up visits, perceptions of mother’s adherence to treatment and follow-up and acceptance of outpatient treatment, and health provider perspectives on whether such an approach should be scaled up in Malawi. Topic guides for mothers of sick young infants addressed decision-making around seeking care, reasons for refusing referral, experiences of outpatient treatment and follow-up, and satisfaction with treatment. Qualitative interviews were conducted by TYIIN study staff and DHMT members in September 2017, towards the end of the study period. The interviews were conducted in Chichewa and detailed notes on responses were recorded in English. The sample size for outcome measurement was based on the primary outcome indicator. Our study assumption was that outpatient treatment completion with the simplified antibiotic regimen through first-level facilities in Malawi would be ≥ 80%. Considering a significance level of 5%, a power of 80% and a one-sided test, we calculated we would need 368 infants with CSI or only fast breathing in the study to demonstrate that treatment completion is greater than or equal to 80%. We conducted the quantitative data analysis using Stata IC 14.2 [StataCorp LP, Texas, USA]. Frequency tables were generated for the primary and secondary outcome variables. We present results separately by case type (CSI or only fast breathing for all variables and critical illness for initial assessment and outcome assessment on day 14). Qualitative data were analyzed using descriptive thematic analysis. Field notes from interviews were reviewed and categorized by the investigators to identify patterns and common themes according to the domains of acceptability and feasibility of outpatient treatment explored during the discussions. The study received ethical approval by the Malawi National Health Science Research Committee (NHSRC) on September 8, 2015. Written informed consent was obtained from all eligible caregivers prior to their participation in the study. To ensure confidentiality and anonymity during analysis, no personal identifiers were entered into the study database and only study personnel had access to the data.