Is the WHO Guide on Essential Practice of Postpartum Newborn Care Used in a District Health Care Facility?

Transcription

1 Journal of Tropical Pediatrics, 2016, 62, doi: /tropej/fmw010 Advance Access Publication Date: 17 June 2016 Original paper Is the WHO Guide on Essential Practice of Postpartum Newborn Care Used in a District Health Care Facility? by Reinhard Klinkott, 1,2 Venance Mushi, 3 Gaudens Komba, 4 Carsten Kr uger, 5 Andreas Schultz, 6 August Stich, 7 Ansgar St ufe, 3 Martin Weber, 8 and Stephan Ehrhardt 9 1 Department of Paediatrics, St. Joseph s Mission Hospital, Peramiho, Ruvuma District, P.O.Box 19 Peramiho Tanzania 2 Department of Paediatrics, Klinikum Garmisch-Partenkirchen, Garmisch-Partenkirchen 82467, Germany 3 St. Joseph s Mission Hospital, Hospital Management, Peramiho, Ruvuma District, P.O.Box 19 Peramiho Tanzania 4 Department of Obstetrics & Gynaecology, St. Joseph s Mission Hospital, Peramiho, Ruvuma District, P.O.Box 19 Peramiho Tanzania 5 Department of Paediatrics, St. Franziskus Hospital, Ahlen 59227, Germany 6 Centrum F ur Internationale Migration Und Entwicklung, Gesellschaft F ur Internationale Zusammenarbeit, Eschborn 65760, Germany 7 Department of Tropical Medicine, Medical Mission Hospital, Wuerzburg 97074, Germany 8 Department of Child and Adolescent Health, World Health Organization, Copenhagen 2100, Denmark 9 Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205, USA Correspondence: Dr. Reinhard Klinkott MD, Auenstr. 6, Garmisch-Partenkirchen. Tel: <Reinhard.Klinkott@klinikum-gap.de> ABSTRACT Background: It is unclear whether algorithms with evidence-based interventions are used in a setting where the work load is high and qualified staff is scarce to identify neonates with life-threatening conditions. Methods: The nurse-midwives knowledge and opinion about the World Health Organization (WHO) guide on postpartum newborn care were assessed in a hospital in Tanzania before and after training. Their adherence to this guide was evaluated by analysing 100 neonatal records. Results: Before training, 44% of the nurse-midwives were familiar with the WHO guide. All nursemidwives supported the implementation of the guide. In all, 21% of the postpartum record forms were fully completed. Risk factors for illness were missed in 27%. Conclusion: Nurse-midwives expertise in the WHO guide on postpartum newborn care cannot be taken for granted. The complexity of this guide demands expertise and makes its use time-consuming and thus its practicability disputable in a setting with limited resources. KEYWORDS: algorithms, delivery of health care, midwifery, newborn, postpartum period, quality of health care INTRODUCTION In 2014, approximately 5.9 million children died worldwide before their fifth birthday [1]. An estimated 2.7 million deaths (45% of all under-five deaths) occurred in the first 28 days of life (neonatal period), with the highest proportion in the first 24 h after birth (up to 37% of neonatal deaths). Ninetynine per cent of these neonatal deaths occurred in VC The Author [2016]. Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oup.com 436

2 Use of WHO Guide in Health Care Facility 437 low- and middle-income countries, mainly in southcentral Asia and sub-saharan Africa [2]. Despite impressive progress on a global scale, the goal to reduce the under-five mortality rate by two-thirds between 1990 and 2015 (UN Millennium Development Goal No. 4) is far from being reached owing to the persistently high neonatal mortality in many of these countries [3, 4]. Evidence-based and cost-effective interventions have been identified to reduce neonatal mortality [5, 6]. These interventions may lead not only to a reduction of up to 50% of all-cause neonatal mortality but may also have an impact on older children, as long-term disability can be averted. However, before these interventions can be implemented, evidencebased algorithms and skilled personnel are needed to identify newborns with potentially life-threatening conditions or at risk for developing serious disease. The Young Infant Study of the World Health Organization (WHO) developed an algorithm that includes seven signs and symptoms (history of difficulty feeding, movement only when stimulated, temperature <35.5 C, temperature 37.5 C, respiratory rate >60 per minute, severe chest indrawing and history of convulsions) to predict the need for hospitalization [7]. These signs and symptoms are included in the algorithm outlined in the WHO guide for essential practice of postpartum and newborn care. This WHO guide aims to reduce the perinatal mortality and morbidity rates in many parts of the developing world [8]. However, its usefulness has never been evaluated. METHODS Study design For our study, we designed a hospital-based prospective observational study at a large rural hospital in Tanzania. We assessed the knowledge and the opinions of nurse-midwives in postpartum newborn care before and after they were taught the essentials outlined by the WHO guide on essential practice of postpartum and newborn care. We also evaluated the adherence to an adapted version of the existing WHO guide used by nurse-midwives with limited neonatal training as part of their daily routine work. Study setting St. Joseph s Mission Hospital Peramiho, a general district hospital, is located in the Ruvuma Region, south-west Tanzania. About 2800 newborns are born every year in the obstetric department with an early neonatal mortality rate (0 6 days of life) of about 10 per 1000 live births [9]. Newborns are primarily seen by nurse-midwives who work in the obstetric department (13 nurse-midwives) and at the reproductive and child health clinic (five nurse-midwives). Selection of study participants Nurse-midwives All nursing staff working in the obstetric department of St. Joseph s Mission Hospital participated in the study. They were registered nurse-midwives who were trained as both nurses and midwives and were the primary providers of care for the newborns. The nurse-midwives were informed that their participation in the study was voluntary. Nurse-midwives who were not willing to follow the study procedure were not disadvantaged. Newborns One hundred and fifty-four newborns were included in the study who were born at St. Joseph s Mission Hospital Peramiho in October/November 2011 and had a birth weight of 1500 g. Neonates were excluded if they presented with an obvious lethal congenital malformation (e.g. anencephaly) or if they required ongoing treatment for >10 min after birth (e.g. ongoing resuscitation). Additional exclusion criteria were inability of the mother to give informed consent (e.g. mentally retarded), refusal of providing informed consent or consent withdrawal. Procedures Nurse-midwives All nurse-midwives of the obstetric department were trained in a 2 h lecture on how to perform the study procedures and how to assess the newborns by using an adapted version of the WHO guide on essential practice of postpartum newborn care (supplemen tary Material 1). A PowerPoint presentation was used and a copy of the pages J1-J12 of the WHO

3 438 Use of WHO Guide in Health Care Facility guide as well as the adapted version was distributed among the participating staff. Before and after the training and right after the study period, the nurse-midwives were interviewed anonymously to explore their knowledge (seven multiple choice questions) and ideas (two multiple choice questions, one open question) about the WHO guide and its use. For this purpose, one questionnaire form was designed. Before its implementation, the form was pretested in the hospital with medical staff not caring for newborns to assess its accuracy (Supplementary Material 2). From the second week of the study onwards all nurse-midwives and four supervising nurse-midwives received a small incentive of 500 Tanzanian shilling (ca US dollars) for each consent form obtained and for each newborn recruited, if the study newborn was also seen by the paediatrician at discharge. No incentives were paid for completing the assessment forms to avoid any impact on the study outcome. Newborns If not admitted for treatment or if not deceased of a serious complication/illness, each study newborn was assessed six times after birth according to the WHO guide of postpartum newborn care: 15, 30, 45 and 60 min after the delivery of the placenta, after the first breastfeeding and before discharge. The assessment was documented on a postpartum record form in English and Kiswahili (Supplementary Material 1). Newborns who were excluded from the study were also assessed according to the same WHO guide used as hospital standard for assessing newborns at that time. The study procedure was pretested on 16 newborns. Their data were not included in the analysis. The first five assessments were performed by the nurse-midwife who conducted the delivery. The sixth assessment was performed first by a nurse-midwife of the reproductive and child health clinic and immediately afterwards by a consultant paediatrician/ neonatologist, unaware of the outcome of the nursemidwife s assessment. On the basis of their findings they had to decide independently whether the newborn needed to be referred to the paediatric department for further management, the newborn needed to be reassessed or the newborn was well and ready for discharge. If a newborn was classified as being ill at any time during the assessment, the newborn was immediately referred to the paediatric department for further management. Statistical analyses The study comprised a convenience sample of 100 newborns assessed by 18 nurse-midwives. Excel was used for data analysis to evaluate the expertise and opinion of the nurse-midwives. The Kappa statistic (defined as the fraction of the observed agreement not due to chance divided by the maximum nonchance agreement) was used to compare the outcome of routine neonatal assessments at discharge as judged by trained nurse-midwives with the outcome of routine neonatal assessments as judged by a consultant paediatrician and neonatologist. The consultant paediatrician/neonatologist was deemed to be the reference standard. Ethical considerations This study was conducted in accordance with the ethical principles of the Declaration of Helsinki [10]. Ethical clearance was granted by the National Health Research Ethics Review Committee of the National Institute for Medical Research in Tanzania (NIMR Act No. 23 of 1979, PART III Section 10(2)) and the hospital management. RESULTS Knowledge and opinions of the nurse-midwives All 18 participating nurse-midwives were trained on essential practice of postpartum newborn care. A total of 18 (100%) nurse-midwives completed the questionnaire before the training, 17 (94%) of these immediately after the training, and 13 (72%) right after the study. Therefore, the results of the respective questionnaire refer to the number of nursemidwives participating in the questionnaire at that time and not to all nurse-midwives participating in the study. In the first questionnaire before the training, 8 (44%) nurse-midwives stated that they knew the WHO guide, 5 (28%) nurse-midwives denied and 5

4 Use of WHO Guide in Health Care Facility 439 Table 1. Nurse-midwives opinion on the WHO guide on postpartum newborn care Question Time of interview Number of Yes nurse-midwives interviewed I do not know No Other No answer or answer not clear WHO guide as part of the daily routine activity? a Before the training (94%) NA NA NA 1 (6%) After the training (88%) 1 (6%) 1 (6%) NA NA Following the study (100%) NA NA NA NA Enough time to use Before the training 18 9 (50%) 4 (22%) 5 (28%) NA NA WHO guide? a After the training 17 3 (35%) NA 10 (59%) NA NA Following the study 13 8 (62%) NA 5 (38%) NA NA Question Time of interview No. of nursemidwives interviewed Opinion on the WHO guide and its use? b (28%) of them were not sure. In the second questionnaire after the training and third questionnaire right after the study, all participating nurse-midwives (100%) noted that they knew the WHO guide. Answers given by the nurse-midwives to the questions on the objectives of the WHO guide on postpartum newborn care in the questionnaire were 68% correct before the training, 94% correct after the training and 94% correct following the study. The majority of nurse-midwives supported the idea to implement the WHO guide on postpartum newborn care as part of their daily routine activity (94% before and 88% after the training, 100% following the study). However, a smaller proportion of nurse-midwives thought that they would have enough time to assess the newborns according to the WHO guide (Table 1). When asked about their opinion on the WHO guide and its use, a majority of the nurse-midwives (53%) appreciated the WHO algorithm in the second questionnaire after the training. It is remarkable that there were even more positive comments (85%) about the WHO guide in the third questionnaire right after the study (Supplementary Material 3). After the training, some of the nurse-midwives (59%) argued more staff would be needed to implement the WHO guide successfully. Yet, this In favour of the guide More staff needed Training is important Other No answer Before the training 18 7 (39%) 2 (11%) 3 (17%) 1 (6%) 6 (33%) After the training 17 9 (53%) 10 (59%) 1 (6%) NA 2 (12%) Following the study (85%) 5 (38%) 1 (8%) NA 0 (0%) a Multiple choice question, only one answer to be marked. b Open question, more than one answer possible. NA ¼ not applicable. proportion decreased after study completion (38%). Only one midwife called for more training. Adherence to the WHO guide Of the 154 included newborns, 28 did not receive a postpartum record form, 1 died at the age of 11 h, 10 were lost to follow-up and 16 were analysed to pretest the postpartum record form. Thus, the assessment data of 100 newborns were finally analysed. Twenty-one of the 100 postpartum records were fully completed by the nurse-midwives. In 61%, the first five assessments after birth were done by just three nurse-midwives (Table 2). The first assessment in the delivery room was completely documented in 91% of the participating newborns. In two newborns, the assessments were not documented. The fifth assessment after the first breastfeeding (on average 2 h after delivery) was least documented mainly in the section look, listen, feel where the performance of kangaroo care and 15 signs of illness had to be checked. Information was only available for 39% of children. The last, sixth assessment at discharge (on average 8.36 h after delivery) with 27 items was completely documented in 78% of participants. Information was missing mainly for the date and

5 440 Use of WHO Guide in Health Care Facility Table 2. Completed items of postpartum record forms and time needed for assessment Number of patient seen (one patient died at the age of 11 h) 100 (100%) 100 (100%) 100 (100%) 100 (100%) 100 (100%) 99 (100%) Time of assessment 15 min 30 min 45 min 60 min After first At discharge After the delivery of the placenta breast-feeding Time of assessment noted 96 (96%) 96 (96%) 95 (95%) 95 (95%) 60 (60%) 91 (92%) Period of time after delivery ND ND ND ND 2:00 hr 32:26 hr of the newborn in average ASK, CHECK RECORD Items to check 8 NA NA NA 1 5 All items completed 97 (97%) NA NA NA 88 (88%) 98 (99%) Items completed correctly 71 (71%) NA NA NA NA NA Not identified by midwife Preterm delivery NA NA NA NA NA NA Mother treated with antibiotics within 2 days of delivery NA NA NA NA NA NA Premature rupture of membranes NA NA NA NA NA NA Difficult birth NA NA NA NA NA NA HIV-positive mother NA NA NA NA NA NA VDRL-positive mother NA NA NA NA NA NA LOOK, LISTEN, FEEL Items to check All items completed 94 (94%) 82 (82%) 80 (80%) 81 (81%) 69 (69%) 99 (100%) CLASSIFY Items to check All items completed 98 (98%) 91 (91%) 91 (91%) 92 (92%) 88 (88%) 99 (100%) Items completed correctly considering the findings in the section ASK, CHECK RECORD or LOOK, LISTEN, FEEL 85 (85%) 83 (83%) 80 (80%) 81 (81%) 70 (70%) 96 (97%) Signature of assessing person 98 (98%) 94 (94%) 94 (94%) 94 (94%) 92 (92%) 99 (100%) Nurse-midwife 1 21 (21%) 21 (21%) 21 (21%) 21 (21%) 20 (20%) 1 (1%) Nurse-midwife 2 2 (2%) 1 (1%) 1 (1%) 1 (1%) 1 (1%) 0 (0%) (continued)

6 Use of WHO Guide in Health Care Facility 441 Table 2. Continued Number of patient seen (one patient died at the age of 11 h) 100 (100%) 100 (100%) 100 (100%) 100 (100%) 100 (100%) 99 (100%) Nurse-midwife 3 4 (4%) 4 (4%) 4 (4%) 4 (4%) 4 (4%) 0 (0%) Nurse-midwife 4 a 21 (21%) 21 (21%) 21 (21%) 21 (21%) 21 (21%) 10 (10%) Nurse-midwife 5 1 (1%) 1 (1%) 1 (1%) 1 (1%) 1 (1%) 0 (0%) Nurse-midwife 6 2 (2%) 2 (2%) 2 (2%) 2 (2%) 2 (2%) 0 (0%) Nurse-midwife 7 a 8 (8%) 8 (8%) 8 (8%) 8 (8%) 7 (7%) 6 (6%) Nurse-midwife 8 11 (11%) 8 (8%) 8 (8%) 8 (8%) 8 (8%) 0 (0%) Nurse-midwife 9 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) Nurse-midwife 10 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) Nurse-midwife 12 8 (8%) 8 (8%) 8 (8%) 8 (8%) 8 (8%) 3 (3%) Nurse-midwife (20%) 20 (20%) 20 (20%) 20 (20%) 20 (20%) 0 (0%) Nurse-midwife 14 a NA NA NA NA NA 1 (1%) Nurse-midwife 15 a NA NA NA NA NA 53 (52%) Nurse-midwife 16 NA NA NA NA NA 2 (2%) Nurse-midwife 17 NA NA NA NA NA 20 (20%) Nurse-midwife 18 NA NA NA NA NA 4 (4%) All items of postpartum 91 (91%) 74 (74%) 72 (72%) 75 (75%) 39 (39%) 77 (78%) record form completed (including time, signature) Time needed for assessment 2:05 min (2) 2:59 min (1) 0:57 min (1) 0:54 min (2) 4:14 min (1) 6:30 min (94) in average (number of assessments observed) anurse-midwife in charge, nurse-midwives 1 13 working in obstetric department, nurse-midwives working in the reproductive and child health clinic. ND ¼ not detected; NA ¼ not applicable.

7 442 Use of WHO Guide in Health Care Facility Table 3. Kappa statistics (j) when comparing the outcome of the nurse with the outcome of the paediatrician a Assessment item j Agreement number (%) Total number documented ASK, CHECK RECORD Has baby had convulsions (99%) 95 Feeding well NA 95 (98%) 97 Mother not able to breastfeed NA 97 (100%) 97 Mother concerned (98%) 97 LOOK, LISTEN, FEEL Grunting NA 99 (100%) 99 Severe chest indrawings NA 99 (100%) 99 Breathing >60/min (99%) 99 Breathing <30/min NA 99 (100%) 99 Capillary refill time >2 s NA 99 (100%) 99 Body temperature >38 C (99%) 99 Body temperature <36.5 C NA 98 (99%) 99 Convulsions (99%) 99 Floppy or stiff (100%) 99 Movement only when stimulated NA 99 (100%) 99 Eyes swollen and draining pus NA 97 (99%) 98 Umbilicus red or draining pus NA 99 (100%) 99 Skin rash (100%) 99 Bleeding NA 99 (100%) 99 Pallor NA 99 (100%) 99 Yellow skin on face, palms or soles and <1 day old NA 97 (98%) 99 One arm does not move NA 99 (100%) 99 Malformations NA 99 (100%) 99 CLASSIFY Baby needs to be referred/treated (84%) 99 Baby needs to be reassessed (83%) 99 Baby is well (95%) 99 On average NA 94% 99 a Poor agreement: j < 0; slight agreement: j ¼ ; fair agreement: j ¼ ; moderate agreement: j ¼ ; substantial agreement: j ¼ ; almost perfect agreement: j ¼ and percentage of agreement; in most of the items, calculation of the kappa statistics was not possible because of the small number. NA ¼ not applicable. time of the assessment and the body weight. This assessment took 6.30 min on average when done by the nurse-midwife and 4.41 min when done by the paediatrician. There was a substantial agreement between the outcome as judged by the nurse and the outcome as judged by the paediatrician. However, the kappa statistics was possible only in 10 of 25 items because of the small number of participants. Therefore, we also calculated the percentage of agreement (94%, see Table 3). Poor agreement was seen in the part classify where mainly the decision for prophylactic treatment was missed by the nurse in eight well newborns (6 owing to HIV exposure, 1 owing to syphilis exposure, 1 failed treatment of the mother owing to premature rupture of membranes). Risk factors for severe illness that had been already checked by the nurse-midwives during the first

8 Use of WHO Guide in Health Care Facility 443 assessment after birth, were assessed again at discharge by the paediatrician. In 27 newborns, important information like the gestational age, the maternal HIV status or the time of the rupture of membranes were not documented. This delayed adequate diagnostic investigations and prophylactic treatment in the neonates. DISCUSSION Knowledge and opinions of WHO practice about postpartum newborn care The results of the questionnaire before the study show that the majority of the nurse-midwives were not familiar with the WHO guide on postpartum newborn care. Therefore, expertise in WHO guidelines cannot be taken for granted, and capacity building is one challenge to be met. The fact that more nurse-midwives appreciated the WHO guide after the study than before may demonstrate their change of attitude during the course of the study. Adherence to the WHO guide The overall adherence to the adapted version of the WHO guide on postpartum newborn care was poor among the participating nurse-midwives. Sixty-one per cent of the first five assessments after birth were done by just 3 of the 13 nurse-midwives in the labour room. Only 21% of the postpartum record forms were filled in completely for all six assessments. The sub-optimal recording of the second, third, fourth and fifth assessment of the newborns may indicate that newborns were checked incompletely. It is worrying that important information in the maternal history was missed in every fourth neonate, putting the newborn at risk for severe illness like neonatal sepsis or HIV infection. Especially the fifth assessment was poorly recorded, with only 39% completed postpartum record forms. This assessment checks the largest number of items and is therefore more time-consuming. This observation parallels the questionnaire results where the nursemidwives mentioned lack of time and staff as an obstacle to use the WHO guide as part of their daily routine activity. But the complexity of an assessment and its temporal aspects cannot be seen as the only determinants for adherence or nonadherence. The last and most complex assessment at discharge was performed completely in 78% of the newborns. This assessment was done in the reproductive and child health clinic where the nurse-midwife had no other responsibilities and therefore in fact more time for the assessment, but where she supposedly felt being supervised as all newborns were also assessed by the study paediatrician immediately afterwards. This may indicate one bias in this study. It was encouraging to note a substantial agreement in the adjudication of the assessment between the nurse-midwife and the study paediatrician in the final assessment. However, risk factors for illness like maternal HIV or syphilis infection or failed treatment of the mother owing to premature rupture of membranes were missed in all exposed newborns. The success of a neonatal screening tool depends on the effective implementation of the planned intervention and its subsequent adherence. Determinants need to be identified and may be specific to a setting and the health professionals involved. They may be economic, political, organizational and behavioural in nature [11 13]. It was not the objective of our study to investigate all these determinants. Furthermore, our findings represent a single-centre study with a limited number of study participants (newborns and nurse-midwives) without formal randomization and control group. The shortage of qualified and committed midwifery staff is certainly one of the main challenges in a low-resource setting when a new intervention is considered to improve the quality of perinatal care [14]. One-off training, which we conducted before the study, is definitely not enough to facilitate long-term capacity building. Our intention was to communicate knowledge that is necessary to assess the newborns. The results of the questionnaire after the training suggest that we succeeded to a certain extent. Nevertheless, lack of expertise might still be a potential barrier in assessing newborns and may explain the differing performance of the nurse-midwives and their adherence to the WHO guide. Therefore, regular bedside training with subsequent control of learning success is needed. But also specialized knowledge alone does not seem to guarantee the adherence to WHO standard

9 444 Use of WHO Guide in Health Care Facility and thereby the careful assessment of the newborns. Skilled health care professionals need to have the time and the commitment to fulfil their duties. In the questionnaire right after the study, shortage of time was mentioned as an obstacle but not by the majority of the nurse-midwives. We did not measure their motivation but we can at least assume by the number of positive comments that most of the nurse-midwives appreciated the implementation of the WHO guide as part of their daily activity. However, their daily commitment may be also influenced by other factors like the work load in the labour room, the complexity of the assessment guide and a system of close supervision and incentives. In our setting, we have seen a much better performance of nurse-midwives in the reproductive and child health clinic despite the complexity of this final assessment. More time for this assessment, less workload for the dedicated nurse-midwife and the feeling of being supervised by the paediatrician doing the same assessment afterwards next door may explain this observation. But we also have to take into account the design effect of the study, as assessments were done by the nurse-midwives and are probably clustered by them. For implementing guidelines, a multifaceted strategy has been shown to be successful in obstetric care [15]. The same strategy may be needed for neonatal interventions. When assessing the WHO guide for postpartum newborn care, various determinants need to be considered that may influence its effective implementation and the subsequent adherence to it. The complexity of the existing WHO is one of these determinants, as it demands expertise and makes its use time-consuming and thus its practicability disputable in a setting where the work load is high and qualified staff is scarce. A simpler algorithm to assess newborns may be one solution to address this problem. It may miss some of the newborns at risk after birth but may be easier to implement as part of the postpartum routine activity. Promising results were seen in the Young Infant Study where an algorithm with only seven signs and symptoms was used [7]. But this algorithm was developed to predict the need for hospitalization in presumably sick young infants presenting to health facilities. Studies have meanwhile shown good results at community level [16]. However, further research is needed to prove its validation as a screening tool also for usually well young newborns in a hospital setting. Another simple algorithm may be the checklist developed in the safe childbirth checklist programme by the WHO [17]. Its effectiveness in terms of reducing neonatal mortality could not be shown so far and has yet to be proven in a multi-centre randomized controlled trial [18]. CONCLUSION Determinants for the successful implementation of and the subsequent adherence to the WHO guide on postpartum newborn care need to be identified and addressed. These determinants may be different from health facility to health facility, which poses a challenge to the collection and analysis of comparable data for operational research on postpartum newborn care. The complexity of the existing WHO guide on essential practice of postpartum newborn care demands expertise and makes its use timeconsuming. Thus, its practicability is disputable in a setting where the work load is high and qualified staff is scarce. Nurse-midwives expertise in WHO guidelines cannot be taken for granted, and therefore, capacity building is one challenge to be met. The development of a simpler screening tool, which can effectively identify sick newborns or those at risk, is easy to remember and not time-consuming to practise may be one approach to guarantee high-quality performance in postpartum newborn care. But even a simple screening tool will not reduce neonatal mortality unless other determinants for the quality of health care provision are considered. Such other determinants may be a system of ongoing training, supervision/incentives, providing enough staff and impact monitoring. Its impact on the successful implementation of evidence-based interventions should be the subject of future operational research on postpartum newborn care. ACKNOWLEDGEMENTS We thank all the nurses and midwives of the St. Joseph s Mission Hospital for their assistance, Dr Isaya Mvinge for the translation of the study forms into Kiswahili, Mrs Theresa Hierl for

10 Use of WHO Guide in Health Care Facility 445 collecting the data and all the mothers and their newborns for their participation in this study. SUPPLEMENTARY MATERIAL Supplementary data are available at Journal of Tropical Pediatrics online. REFERENCES 1. UNICEF, WHO, The World Bank, United Nations. Levels and Trends in Child Mortality: Report New York, NY: UNICEF, Lawn JE, Cousens S, Zupan J; Lancet Neonatal Survival Steering Team. 4 million neonatal deaths: when? where? why? Lancet 2005;365: Annan K. Road Map Towards the Implementation of the United Nations Millennium Declaration. Report of the Secretary-General. United Nations Document A/56/326. New York: United Nations, Rajaratnam JK, Marcus JR, Flaxman AD, et al. Neonatal, postneonatal, childhood, and under-5 mortality for 187 countries, : a systematic analysis of progress towards Millennium Development Goal 4. Lancet 2010;375: Darmstadt GL, Bhutta ZA, Cousens S, et al.; Lancet Neonatal Survival Steering Team. Evidence-based, costeffective interventions: how many newborn babies can we save? Lancet 2005;365: Bhutta ZA, Das JK, Bahl R, et al.; for The Lancet Newborn Interventions Review Group; The Lancet Every Newborn Study Group. Can available interventions end preventable deaths in mothers, newborn babies, and stillbirths, and at what cost? Lancet 2014;384: The Young Infants Clinical Signs Study Group. Clinical signs that predict severe illness in children under age 2 months: a multicentre study. Lancet 2008;371: World Health Organization. Pregnancy, Childbirth, Postpartum and Newborn Care. A Guide for Essential Practice. Geneva: WHO, St. Joseph s Mission Hospital Peramiho. Annual Report World Medical Association Declaration of Helsinki - Current (2008) version. 11. Rowe AK, de Savigny D, Lanata CF, et al. How can we achieve and maintain high-quality performance of health workers in low-resource settings? Lancet 2005;366: Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patient care. Lancet 2003;362: Althabe F, Bergel E, Cafferata ML, et al. Strategies for improving the quality of health care in maternal and child health in low- and middle-income countries: an overview of systematic reviews. Paediatr Perinat Epidemiol 2008;22(Suppl. 1): Nyamtema AS, Urassa DP, Massawe S, et al. Staffing needs for quality perinatal care in Tanzania. Afr J Reprod Health 2008;12: Chaillet N, Dube E, Dugas M, et al. Evidence-based strategies for implementing guidelines in obstetrics: a systematic review. Obstet Gynecol 2006;108: Darmstadt GL, Baqui AH, Choi Y, et al.; and the Bangladesh Projahnmo-2 (Mirzapur) Study. Validation of a clinical algorithm to identify neonates with severe illness during routine household visits in rural Bangladesh. Arch Dis Child 2011;96: World Health Organization. Safe Childbirth Checklist Programme An Overview. Geneva: WHO, Spector JM, Agrawal P, Kodkany B, et al. Improving quality of care for maternal and newborn health: prospective pilot study of the WHO safe childbirth checklist program. PLoS One 2012;7:e35151