Facility characteristics
All facilities had a designated maternity ward with specific room/s within the ward for deliveries. The delivery rooms had similar physical layouts with designated areas for labour and delivery, newborn assessment, waste disposal, and storage of personal protective equipment (PPE) and birth kits.
All eight labour and delivery wards had functional handwashing facilities with soap, alcohol based hand rub (ABHR) and gloves (clean and sterile). Water was available via a sink with a connected tap in all delivery wards except for one RH which was undergoing construction. All handwashing facilities in the delivery room were visibly clean, accessible and had available and visibly clean hand drying materials. Hand hygiene posters were present, visible and displayed at the handwashing facilities in all but one delivery room.
Structured observations
Participant information
A total of 45 mothers were observed; 22 from the PHC and 23 from the RH. Mothers from the PHC and RH had similar characteristics with a mean age of 28 (21–40) and had an average of 2 (0–6) previous live births. The average travel time to the HCF was 19 minutes (5–40).
The average number of facility birth attendants present per delivery was 2 (1–5). The midwife was the most common birth attendant, present for 100% of all births. 18% of the deliveries were attended to by only one birth attendant.
Labour flow
A total of 95 labour flows were observed, for an average of 2 (range: 0–10) vaginal examinations per woman. Birth attendants initiated only 22% of the labour flows with adequate hand hygiene (Table 3).
Table 3
Hygiene risk categories prior to all flows combined by provider type, facility type and work shift.
| n | Adequate | Inadequate | Aseptic Technique Invalidated | Somers’ D clustered by facility; p-value (Confidence interval) |
Flow type | | | | | |
Labour | 95 | 21 (22%) | 52 (55%) | 22 (23%) | Ref |
Delivery | 102 | 19 (17%) | 36 (35%) | 47 (46%) | 0.25; p = 0.00 (0.15 – 0.35)*** |
Newborn aftercare | 54 | 4 (7.4%) | 11 (20.4%) | 39 (72%) | 0.46; p = 0.00 (0.34–0.59)*** |
All flows | 251 | 44 (18%) | 99 (39%) | 108 (43%) | |
Provider type | | | | | |
Sec. Midwife | 145 | 27 (19%) | 54 (37%) | 64 (44%) | Ref |
Primary Midwife | 93 | 15 (16%) | 39 (42%) | 39 (42%) | -0.05; p = 0.8 (-0.34–0.25) |
Intern | 3 | 1 (33%) | 1 (33%) | 1 (33%) | 0.02; p = 0.14 (-0.01–0.04) |
Doctor + Nurse | 5 | 0 (0%) | 3 (60%) | 2 (40%) | -0.05; p = 0.206 (-0.12–0.03) - |
Facility type | | | | | |
Primary Health Centre | 137 | 17 (13%) | 58 (42%) | 62 (45%) | Ref |
Referral Hospital | 110 | 26 (24%) | 39 (35%) | 45 (41%) | 0.24; p = 0.20 (-0.12–0.60) |
Shift | | | | | |
Morning | 109 | 24 (22%) | 35 (32%) | 50 (46%) | Ref |
Afternoon | 49 | 10 (20%) | 18 (37%) | 21 (43%) | -0.03; p = 0.59 (-0.16 − 0.09) |
Overnight | 89 | 9 (10%) | 44 (49%) | 36 (41%) | -0.25; p = 0.03 (-0.47 – -0.02)* |
Over half of observed labour flows (55%) were initiated under inadequate hand hygiene and 23% initiated when aseptic technique had been invalidated.
Delivery flow
Birth attendants initiated the majority (46%) of the delivery flows (n = 102) when aseptic technique had been invalidated, 35% when hand hygiene was inadequate and only 19% under adequate hand hygiene (Table 3).
The proportion of flows that maintained, dropped or improved a hand hygiene category by the end of the delivery flow is represented graphically in Fig. 1. Only 7 of the 19 (37%) flows initiated under adequate hand hygiene maintained this status throughout the delivery flow; 5 of 19 (26%) dropped to inadequate hand hygiene, all of which were glove changes without intermediary handwashing during manual removal of placenta procedures.
7 of 19 of flows (37%) that were initiated under adequate hand hygiene had dropped to invalidated aseptic technique at the end of the delivery flow. The most common reason for invalidation of aseptic technique was donning other PPE items such as non-sterile apron and boots after having already conducted adequate hand hygiene.
The majority of delivery flows that started under inadequate hand hygiene (69%) and those where aseptic technique had been invalidated (83%) maintained those categories throughout the flow. Wiping off faecal and bloody matter from the perineum, floor and trolleys between procedures without subsequent hand hygiene action was the most common observed activity for invalidated aseptic technique. There were only limited improvements in hand hygiene during delivery flows, however these improvements were all inadequate (only gloves changed) and did not fully adhere to hand hygiene protocol.
Newborn aftercare flow
72.5% of all newborn aftercare flows (N = 54) were initiated when aseptic technique was invalidated, 20.5 % with inadequate hand hygiene and 7% with adequate hand hygiene (Table 3). All the newborn care flows initiated under adequate or inadequate hand hygiene maintained these categories throughout (Fig. 2).
Only 13% (5/39) of newborn aftercare flows that were initiated when aseptic technique was invalidated improved over the course of the aftercare flow. All observed improvements in hand hygiene occurred just prior to immunisation of the baby when gloves would be donned without handwashing with soap.
Invalidation of hygiene protocol was more likely as the birth process progressed (Table 3). Both delivery and newborn aftercare flows were more likely to be initiated with invalidated hand hygiene compared to labour flows [Delivery: Somers’ D = 0.25, p = 0.00; Newborn: Somers’ D = 0.46, p = 0.00] and invalidated hygiene practices also more likely to be initiated under newborn aftercare flows than delivery flows [Newborn: Somers’ D = 0.23; p = 0.00; VE: Somers’ D = -0.24, p = 0.00, data not shown]. When all flows are combined, compliance did not differ by facility type (RH vs PHC) or birth attendant qualification and differed significantly by working shifts. Compared to the morning shift (6:00–12:00), adequate hand hygiene was less likely to be practiced during the overnight shift (18:00–6:00) [Overnight: Somers’ D = -0.25; p = 0.03] but did not differ significantly with the afternoon shift (12:00–18:00) [Afternoon: Somers’ D = -0.03; p = 0.59].
Qualitative Results
Participant information
Qualitative data was collected from 4 HCF; 3 PHC and 1 RH. A total of 20 key healthcare workers were interviewed across the HCF. Not all participants completed all data collection activities and the exact activities depended on the specific respondent or the time available for the interviews.
Interviews with midwives revealed insights into behavioural determinants that promoted or constrained the performance of adequate hand hygiene practices. The findings below are organised according to the key components of the BCD approach (Table 1). Within each component, the identified relevant key behavioural determinants are summarised.
Brain + Body
Knowledge and risk perception
Overall, midwives demonstrated adequate awareness and general understanding of the importance of adequate hand hygiene and gloving practice and their associated link to infection transmission. However, knowledge around avoiding glove and clean hand recontamination including the correct hand hygiene protocol following recontamination was limited. For example, midwives correctly identified washing hands with soap and wearing new gloves before conducting vaginal examinations during labour as proper protocol but did not consider it necessary to do so when they used their gloved hands to wipe a blood spot off a surface during delivery.
Infection risk perception around newborn aftercare was low. With the exception of cord care, midwives considered all other newborn aftercare events in the delivery room including the initial assessment of the newborn following birth, as very low infection transmission risk events. This corresponded to the structured observations where gloves were almost always changed prior to cutting the cord but very few hand hygiene actions observed around other newborn aftercare events.
Refresher midwifery training was irregular and usually externally held with limited opportunities for attendance. Typically, only one midwife and facility director were facilitated to attend external trainings. The extent to which hand hygiene was covered varied by training with some being hand hygiene specific while others included hand hygiene within larger training on general maternal and child health. Most of the midwives could not remember the last time they had attended a training and recalled their formal midwifery education program as the last time anyone had provided formal hand hygiene and IPC training. Midwives relied on knowledge sharing meetings held by the few HCWs who attended the external formal trainings and informally through observing and following peers’ practices, particularly those of the senior midwives, during deliveries.
Senses and motives
Midwives pointed to the visibility of potential contaminants serving as their cue for reactive hand hygiene. The presence or contact with soiling, particularly faecal matter, was considered disgusting by all midwives, triggering hand washing with soap and glove changes. However, in most of the structured observations, the presence of visible soiling would typically trigger cleaning/wiping actions such as wiping of soiling on the perineum, delivery surface or floor with, further contaminating the gloved hands, but no subsequent hand or glove hygiene actions would be practiced.
Discounts
Midwives often reported leaving out or forgetting hand hygiene steps particularly when they were under high pressure situations such as birth complications, quick labours, solo shifts and multiple women in labour. Shortage of staff particularly during the night shift was also a commonly reported challenge. In these situations, forgetting to wash hands with soap and change gloves in between or using the same pair of gloves from start to finish was considered common, although in other cases midwives employed these as deliberately as time saving practices. Multiple gloving was also considered a common and acceptable time saving practice.
Behaviour settings
Roles and responsibilities: Across all HCF, midwives had a strong sense of ownership towards the delivery room and assumed all the responsibility for all activities that took place in the room including its general appearance. In addition to performing deliveries, midwives’ responsibilities were to: ensure the availability of clean and sterilised PPE, delivery equipment and hand hygiene materials for each delivery; ensure that hand hygiene and glove use protocols are followed during delivery; maintain a clean and odourless delivery room after each delivery; and, ensure regular cleaning, sterilising and proper waste management.
Prior to each delivery, the roles were decided by the midwives, with one midwife primarily responsible for care of the mother during labour and delivery including cutting the cord, and the another midwife taking on newborn aftercare such as physical inspection of the baby, taking weight and measurements and supporting the breastfeeding initiation. These roles switched with each delivery depending on discussion and agreement with the midwives available on shift. All birth attendants present in the room assisted and supported each other throughout the delivery process.
Props and infrastructure
Similar to findings from the structured observations, midwives reported the regular availability of functioning and accessible handwashing facilities. In the rare event of a lack of water supply, midwives reported tasking the woman’s relatives to bring enough water into the delivery room for the duration of the delivery. Similarly, running out of gloves, handwashing soap and alcohol rub was not common. The stock-out reporting process was simple and well understood, and midwives were responsible for monitoring the stock of these items daily and reporting to the facility director or accountant directly or during the regular staff meetings whenever supplies were low.
When an unanticipated stock out did occur, midwives reported compromising hygiene practices in the interim such as using non-sterile gloves to carry out procedures until sterile gloves were replaced or limiting the frequency of glove changes per delivery. In other cases, midwives bought missing materials using their own money and were reimbursed later.
Routine
Multiple gloving and the subsequent layer-by-layer removal during the delivery flow was the accepted standard of practice in facilities and integrated into the standard caregiving routines. Wearing only one pair of gloves was unanimously considered a serious infection risk for the midwife and mother, and a midwife would only likely do this if there was a glove shortage. Midwives typically wore two or more pairs of gloves at the beginning of delivery consisting of one pair of clean gloves and one or more pairs of sterile gloves. ‘Changing gloves’ was described by the midwives as removing the top pair and either immediately proceeding with the delivery process with the gloves underneath or donning an additional layer of new sterile gloves prior to proceeding. In line with direct observations, midwives reported routinely removing their outermost gloves at two time points; prior to cord cutting and before delivery of the placenta. Handwashing with soap was never conducted in between these removals as the midwife considered the gloves underneath to still be sterile.
Environment
Social environment
Within the delivery room, the midwives, superseded only by doctors, were typically at the top of the social hierarchy. In this setting, midwives were highly respected, listened to and were considered authoritative figures by the nurses, patients and the visitors. The social environment between the midwives was cohesive and reported to be generally very supportive and with mutual respect for each other regardless of rank. Midwives perceived their hand hygiene behaviours as easier to change compared to those of the cleaners and visitors because of their strong social relationship crediting the presence of strong systems of support, knowledge sharing and accountability among each other, with everyone being open to correction, and willing to follow and learn from one another.
Maintenance of the social standing of a fellow midwife was typically prioritised over the immediate danger of potential infection transmission to the mother. For the majority of midwives, correction of a colleague in front of a patient was considered an upset to that social order. Midwives also perceived public correction to lead to loss of trust/confidence and respect between the mother and the HCW. As a result, when a breach of hygiene protocol was observed during the delivery process, a midwife would wait for a more private time to point this out over immediate real-time correction. Conversely, because of this hierarchy, midwives felt no discomfort or hesitation in immediately correcting any visitors’ behaviours when noncompliance to hygiene practices was observed, publicly or otherwise.