Table 1 Study characteristics

Country; site

Nigeria; Lagos

Nigeria; Ebonyi and Kogi

Iran; Lorestan

Ghana; North & South Birim Districts

Zimbabwe

India; Uttar Pradesh

Nigeria; Edo State

Zanzibar, Tanzania

Thai-Myanmar border; Mae La refugee camp

Cambodia, Lao PDR, Mongolia, Papua New Guinea, Philippines, Solomon Islands, Viet Nam

Vietnam; Ho Chi Minh City

Iran; Kurdistan

India; Karnataka

India;

Telangana, Andhra Pradesh

Ghana; Accra

Study design

Cross-sectional

Cross-sectional

Cross-sectional

Cross-sectional

Cross-sectional

Repeated Cross-sectional (nested in randomised trial)

Cross-sectional

Cross-sectional

Cross-sectional

Cross-sectional

Pre-post multi-component intervention

Cross-sectional

Pre-post multi-component intervention

Cross-sectional

Cross-sectional

Facility type

1 secondary and 1 tertiary maternity care facility

2 Primary healthcare facilities

2 secondary healthcare facilities

2 Tertiary healthcare facilities

9 public hospitals

1 public hospital, 6 public health posts,

5 private maternity homes

2 University of Zimbabwe Central Hospitals i.e. National referral hospitals

15 healthcare facilities of the 60 selected for the intervention. The 60 facilities varied between primary and community health centres and first level referral units.

63 healthcare facilities including primary health centers, private clinics, two secondary/district hospital, 2 tertiary/teaching hospitals

1 referral hospital

1 maternity Hospital

3 Cottage Hospitals

1 private Hospital

3 district Hospitals

1 Primary healthcare Unit

Shoklo Malaria Research Unit Clinic

76 first level referred hospital

25 tertiary hospitals

Hung Vuong University Hospital

Be-Sat Hospital of Sanandaj and Hafte-Teer Hospital of Beejar

Sub-district level hospital (basic emergency obstetric care and C-sections)

26 Public secondary healthcare facilities

4 public tertiary healthcare facilities

5 private tertiary healthcare facilities

Korle-Bu Teaching Hospital (tertiary healthcare facility)

Unit/ward

Maternity ward

Labour ward

Unclear. Presumably labour ward

Unclear. Presumably labour ward

Labour & postnatal ward

Labour wards

Delivery wards

Labour wards

Birth centre

Delivery rooms

Delivery suite

Labour & delivery wards

Unclear. Presumably labour ward

Labour ward

Emergency Room and Labour ward

Effect size

None

None

None

None

None

None

None

None

None

None

None

None

None

None

None

Intervention

None

None

None

None

None

Introduction of Safe Childbirth Checklist with peer coaching

None

None

None

None

Yes; educational intervention

None

Yes; testing checklist

The study is part of a baseline evaluation of a quality improvem. Intervention

None

Health professionals involved

Midwives

Doctors, midwives, auxiliary staff

Unclear. Midwives are mentioned in the discussion

Midwives, midwives’ assistants and lay women trained by midwives

Midwives

Any birth attendants

Attending midwives

All staff involved in assisting deliveries

Literate skilled birth attendants resident in the camp and trained by the clinic (not previously trained in midwifery)

Unclear. Any birth attendants

All healthcare workers in the delivery suite. Across all departments in the study they capture doctors, nurses, midwives and technicians^a

Unclear

Any healthcare worker^a (nurses & obstetricians) who cared for women and newborns from admission for childbirth to discharge

Health care providers working in the labour ward

Doctors and nurses

Type of patient-attendant interactions

52 women during delivery and immediate postpartum

31 women in active labour (cervical dilation> 3) and admitted to delivery

200 (low risk) pregnant women^b

18 vaginal deliveries and 22 neonatal cord-care events

20 observations in the labour and 17 in the postnatal wards

1277 deliveries.

Specifically before pushing and soon after birth.

Unclear. Mentions examination and procedures requiring gloves

781 aseptic procedures during labour and delivery

20 births

371 deliveries

All types of hand hygiene opportunity in the delivery suite

96 women with low risk pregnancies^b

405 vaginal examinations at admission and 388 deliveries

242 pre-vaginal examination and 235 deliveries

Unclear

Observation period

May 2010

4 weeks in July 2017

Unclear

2 Months. August–September 1991

May to June 2014

6 to 12 weeks during the intervention from Dec 2014 to Sep 2016

January to May 2011

November 2015–April 2017

6 weeks. Nov-Dec 2008

Unclear

2016–2017

August 2014–May 2015

Throughout 2006

Baseline: Jul-Sept 2010; Endline: Sept.-Dec 2010

May 2016–August 2016

3 weeks. September 2011

Data collectors

Unclear

Qualified midwifes

Unclear

Project director and co-director (a Ghanaian nurse)

3 midwives researchers. 2 working at the study institution. 1 just left the study institution

Trained nurses

Unclear. Trained staff

3 Trained midwifes for each labour ward

2 Dutch midwifery students (4th year)

Trained doctors, nurses, midwives or public health professionals

6 infection control staff trained in direct observation. Unclear if worked in study institute

Unclear.

“Researcher”

Student nurses previously unknown to hospital staff with no clinical responsibility

12 nursing graduates trained for direct observation

6 nurses specifically trained in infection control

Tool used for observation

Checklist developed for study. Based on protocol by Christensson et al. (2001) [33]

Standardised direct observation tool developed for study. Based on a previous tool developed for a qualitative care study. But it references a report [34]; unclear

Checklist developed for study. Content validity assessed

Checklist created for study using criteria from e.g. the WHO Global Programme on AIDS, 1989

Checklist for labour ward developed for study

The WHO Safe Childbirth Checklist

Adapted tool from a previous study based in India. Unclear reference

Observation tool developed for study based on WHO guidelines on hand hygiene in healthcare, 2009 and WHO hand hygiene technical reference manual, 2009

Checklist developed for study, drawing on WHO Safe Motherhood Needs Assessment v1.1 2001

Standard checklist based on EENC Module 1: Annual Implementation Review and Planning Guide

Checklist using the WHO Guidelines on Hand Hygiene in Health Care, 2009. Observation checklist content validity reviewed by MoH and University staff

Tool developed for study based mainly on WHO’s protocol of normal birth, 1997& 2006

WHO Safe Childbirth Checklist presumed to have been used

Checklist adopted from

the WHO concept of five moments of hand hygiene, 2010

Modified version of the WHO form for hand hygiene direct observation, 2010

Study aim disclosed to participants

Unclear. Non participant observation

Study aims were explained to the participants and to the whole staff

Unclear

Participants not told when observation would take place or what practices were observed

Observers were “inside participants” assisting midwives in their work. Checklist was filled after procedures in private

Birth attendants were aware of the observation –observation included many aspects of quality of care; not hand hygiene only.

Unclear. Birth attendants were not previously informed of the walk-in visits but they were not blinded.

Healthcare workers were aware of the observation but were told that the observation was about overall quality of care (not specifically hand hygiene)

Unclear

Unclear

Unclear. Healthcare workers were aware of the observation period

Study aims were explained to the participants midwives

Nature of the intervention, which included awareness practices included in the WHO Safe Birth Checklist (e.g. hand hygiene), presumably clear to participants

Unclear

Unclear. Health workers in these service centres were not aware of being observed

Sampling

Unclear

The facilities with the highest number of deliveries were selected for each state (one primary, one secondary and one tertiary). Any woman who met inclusion criteria and gave consent was invited to participate up to 5 woman per facility. Unclear how the timing of facility visits was scheduled

Non-random quota sampling used to recruit 200 women. 10–30 selected in different stages of labour in each hospital. Sample size calculations justified. Unclear how different stages of labour or women and timing of visits were selected

Observation took place when the project staff visited a facility at a time when a woman was in labour. All midwives on duty when observation took place were included. Occasionally called by facility when delivery expected. Not clear how the timing of facility visits for observation were scheduled

All midwives at the time of the observation were included in the study. Not clear how the timing for facility visits were scheduled

15 facilities for independent observation were based on pragmatic sample. Independent observer visited facilities during non-intervention days for a period of 6 to 12 weeks with the goal to reach 240 pose points in each facility. A mother was observed for as many pause points as possible.

Not clear how the timing for facility visits were scheduled

Public and private health facilities with high caseloads of pregnant women were selected from eight Local government areas. Presumably one walk-in visit for each hospital. Not clear how visits were scheduled and how many deliveries were observed

Observation occured in the 10 highest-volume labour wards for a mode of 6 days each (5–14 days range) for 24 h a day. All attendants. Involved in assisting deliveries during observation.

Not very clear how they selected which healthcare worker to observe.

Sample size calculations justified.

Unclear

Random selection of 3 national or regional tertiary hospital, 4–12 provincial hospital and 2–4 district hospitals in each country.

Deliveries were observed over 1–2 days in each hospital selected. Un clear the selection of delivery observation period; but it mentioned observation was limited to the time of the assessments. Not clear how the timing for facility visits were scheduled

Unclear

Women’s selection – quota sampling (1 in 3 women) proportionally divided between morning, evening and night shifts. Not clear if all women received the full set of observations

Observation took place 24-h for a minimum of 6 days weekly; unobserved days were random. Observation was carried out at admission, from start of pushing to 1 h after birth, discharge. Unclear how women were selected each stage

In the labour room observer spent 6 h a day (either morning or evening shift) for 6 days observing 2–3 mothers a day. No details on how visits or shifts were scheduled. Only one woman was observed at the time

Observation in times & locations with high care density. Each centre was observed at a different time of day for 2 days between 8 AM-5 PM. Not clear how they selected which healthcare worker to observe

Water/Soap/handrub availability

Unclear. Sinks were not located in convenient locations

All facilities had soap and water in the delivery unit but during 1 observation there was no soap. All delivery units had a sink with connected tap available but 2 used veronica buckets. No disposable towerls

None

Unclear. Only reported missing items. Water, soap, handrub were not mentioned as missing

Unclear. They report broadly that basic supplies were often unavailable (not clear is specific to hand hygiene supplies)

Unclear

24-h running water was present in 52% of the observed facilities. Soap in 65% of facilities.

Unclear

Unclear. All essential equipment for standard antenatal care, and essential care of obstetric complications was present.

147 hospitals assessed for WASH services. 72% of hospitals had clean sinks with running water, soap or handrub in the delivery rooms. Data is not specific for the 101 hospital where deliveries were observed

None

None

Unclear. Hospitals selected based on general availability of supplies

unclear

Resources observed once. Water, soap and single-use towel for drying available on labour ward. Handrub not available