Quantifying health facility service readiness for small and sick newborn care: comparing standards-based and WHO level-2 + scoring for 64 hospitals implementing with NEST360 in Kenya, Malawi, Nigeria, and Tanzania

The framework for this approach was developed based on the HSBB framework adapted from WHO with a focus on six HSBBs [17]. The six adapted HSBB modules focus on: 1) infrastructure; 2) medical devices and supplies; 3) human resources; 4) information systems; 5) family-centred care; and 6) governance. Sub-modules for each HSBB were also identified, and are further described in Table 1. Service readiness was assessed according to national and global standards for SSNC. All items required for standards-based service readiness of SSNC, where a clear national or global standard exists, were included in the analysis. Items that did not have a clear national or global standard were excluded from analyses. Items were grouped into HSBB modules and sub-modules to support identification of health systems gaps for action (Table 1, Additional File 1).

HFA data were cleaned, and quality checked in REDCap and during analysis using Stata 17 (StataCorp LLC, Texas, USA). Items were scored on availability (e.g., nurses staffed on the neonatal unit), or if relevant on availability and functionality (e.g., devices). Functionality was defined according to the item. For example, functional devices must be in working order, and functional clinical guidelines must be available and easily accessible at the time of the HFA. No points were given if not available, one point if available, and two points if available and functional. Scores were calculated by HSBB with a percentage based on total score over the total possible score for that HSBB. Items with missing responses were removed from the denominator to be conservative. A sensitivity analysis was performed to determine the impact of removing items with missing responses from scores compared to assigning missing items a score and keeping them in the denominator. Scores were analysed using the neonatal unit as the unit of analysis. At the country level, scores were aggregated using the median and interquartile range of all neonatal unit scores in that country.

Service readiness criteria for assessment of WHO level-2 SSNC and transition to level-3 care were developed through selection of ten interventions according to the WHO levels of care framework for level-2 + SSNC (Additional File 2) [1]. Level-2 + SSNC includes seven level-2 interventions: 1) thermal care; 2) intravenous and assisted feeding; 3) safe oxygen use; 4) provision of antibiotics guided by culture report; 5) jaundice management with phototherapy; 6) management of neonatal encephalopathy; and 7) referral for congenital abnormalities. Level-2 + SSNC also includes three transition to level-3 interventions: 1) CPAP use; 2) exchange transfusion; and 3) follow-up of at-risk newborns. This focus on level-2 + SSNC is aligned with the content of the NEST360/UNICEF HFA tool and the ENAP coverage target for 80% of districts in every country to have at least one functional level-2 + inpatient newborn care unit with respiratory support by 2025, which requires evaluation of level-2 + care to assess progress [4, 18].

Sub-modules for each clinical intervention were scored on a zero to three scale – allocating zero points if "not ready", one point if "basic readiness", two points if "comprehensive readiness", and three points for "gold standard care". Basic, comprehensive, and gold-standard readiness were defined according to national and global clinical guidelines, NEST360 educational materials, and review by clinicians with SSNC experience as described above. These sub-module scores included individual items needed for each individual pathway, however, scores were allocated for the overall sub-module. For sub-modules with missing responses, missing items were removed from the analysis to be conservative and allow all neonatal units to achieve the highest possible score regardless of missing data. Sub-module scores were added to calculate an intervention-level score. Intervention scores were calculated by a percentage based on total sub-module score over the total possible sub-module score for that intervention. Follow-up of at-risk newborns, which only included clinical guidelines, was excluded from the overall score to be conservative. Therefore, nine clinical interventions (thermal care, intravenous and assisted feeding, safe oxygen use, provision of antibiotics guided by culture report, jaundice management with phototherapy, management of neonatal encephalopathy, referral for congenital abnormalities, CPAP use, and exchange transfusion) were set to contribute equally to the neonatal unit total percentage score and were summarised using heatmaps and presented by country and clinical intervention. Heatmaps present scores using five colours to represent each 20% increase in scores, with 0–20% as lowest (dark red) and 80–100% as highest (green). Stacked bar charts present scores demonstrating the number of sub-modules with no readiness (dark red), basic readiness (orange), comprehensive readiness (yellow), and gold-standard readiness (green). Two additional categories were assessed to evaluate other clinical care items and health systems resources needed for the clinical interventions, but these were not included in the overall level-2 + scores, which focused on interventions only. Scores were analysed using the neonatal unit as the unit of analysis. At the country level, scores were aggregated using the median and interquartile range of all neonatal unit scores in that country.