**Background:** Reducing newborn mortality remains a global priority, with most of the 2.76 million annual neonatal deaths occurring in low- and middle-income countries (LMICs). While parental involvement in newborn care improves outcomes such as bonding, breastfeeding, and shorter hospital stays, most evidence comes from high-income countries. In LMICs, nurse-to-baby ratios far exceed international recommendations (e.g., 1:1 for intensive care in the UK vs. 7–15 babies per nurse in Nairobi), leading to informal delegation of care to mothers. This study aimed to describe the roles mothers play in caring for their sick newborns in two contrasting Kenyan NBUs and examine how structural, economic, and social contexts shape participation.
**Methods:** An ethnographic approach was used, with data collected through 627 hours of non-participant observations between March 2017 and August 2018 in one government-funded and one faith-based hospital in Nairobi. Observations covered day and night shifts across weekdays and weekends. A total of 40 discharge interviews (20 per hospital) were conducted with purposively selected mothers. Data were analysed using a modified grounded theory approach informed by critical medical anthropology.
**Key Results:** The two hospitals served markedly different populations. The faith-based hospital, located in an affluent neighbourhood, required a mandatory deposit of ~USD 300 with daily costs of USD 35–150 (vs. a minimum monthly wage of ~USD 188). Most mothers there had post-secondary education, full-time employment, and medical insurance. In contrast, mothers in the government hospital, located next to a slum, were mostly unemployed or in informal work without insurance. Nurse-to-baby ratios differed dramatically: in the government hospital, one nurse often cared for 15–40 newborns per shift, while the faith-based hospital maintained ratios of 1:1 for NICU and up to 1:4 for stable babies.
In the government hospital, mothers became involved in care within the first 24 hours of admission, performing diaper changes, top-tailing (washing), and nasogastric (NG) tube feeding. Nurses provided minimal instruction; mothers learned primarily from other mothers through peer support. For example, one first-time mother (Hellen) learned how to change diapers and feed via NG tube from a neighbouring mother after receiving only brief instructions from a nurse. Mothers also developed practical strategies (e.g., pinching the NG tube to regulate milk flow) and cultural practices (placing paper on the baby's forehead to prevent hiccups). In contrast, in the faith-based hospital, nurses performed all care initially, and mothers only began participating days after admission, after the baby stabilised and moved to the general ward. Mothers received one-on-one teaching from nurses for tasks like bathing.
Breastfeeding support was inadequate in both hospitals. In the government hospital, mothers struggling to express milk were sometimes scolded by staff, with support coming from other mothers. In the faith-based hospital, formula feeding was logistically easier and allowed, though some doctors insisted on breastmilk, creating tension. Mothers in the government hospital appreciated their involvement for enabling bonding, while those in the faith-based hospital reported feeling powerless, with one stating: 'I tend to come in only during visiting hours just to bring milk because even if I stay here, there is nothing much I can do anyway.' Mothers in the faith-based hospital expressed greater concern about post-discharge care without nursing support.
**Clinical Implications:** The study reveals a paradox: mothers in the understaffed government hospital participated more in care but with inadequate training and supervision, risking dangerous mismanagement (e.g., choking during NG feeding). Mothers in the better-staffed faith-based hospital had less involvement initially, leading to reduced confidence and bonding. In both settings, mothers' physical and psychological needs were largely ignored. The authors argue for contextually appropriate interventions that recognise the mother-infant dyad, improve nurse training to support mothers, address chronic understaffing, and promote family-centred care. The recent WHO iKMC trial (showing a 25% reduction in mortality with immediate KMC) offers a model, but requires significant system changes. A first step is equipping nurses to identify and support mothers' needs and encouraging mother-infant bonding within the first 24 hours.