· Web viewSAVING NEWBORN LIVES. South Africa . Good Start III . End of Study Report . September....

Results

SAVING NEWBORN LIVES

South Africa

Good Start III

End of Study Report

September 2012

ACKNOWLEDGEMENTS

We would like to acknowledge all the mothers and infants who participated in the study. We would like to acknowledge the support of Joy Lawn and Deborah Sitrin throughout the course of the study. The work of Gail Snetro and Thoko Ndaba on the facility strengthening was also incredibly valuable.

I. INTRODUCTION

South Africa is one of only a dozen countries worldwide where child mortality has increased since 1990.1 This is primarily related to the HIV epidemic and more than half of the child deaths are attributed to HIV/AIDS.2 Much less attention is given to the fact that 30% of under-five deaths in South Africa are neonatal deaths in the first 28 days of life, a total of around 20,000 deaths each year. 3 Progress towards MDG4 in South Africa will depend largely on scaling up effective prevention of mother to child transmission (PMTCT) of HIV and also addressing neonatal mortality. This imperative drives increasing focus on the neonatal period, including improving facility based care and particularly to develop and test appropriate models of sustainable community-based care in South Africa in order to reach the poorest families, especially the peri-urban poor.

A review of a variety of integrated service delivery packages has previously suggested 8 integrated

MNCH packages which also include PMTCT of HIV. 4 This global set has been adapted for national application in South Africa (see Web Appendix 1). 3 Our study seeks to address gaps in previous research undertaken in South Africa namely a assessing strategies to reach the poorest, still heavily disadvantaged mothers and newborns, echoing wider policy and programme debates regarding the roles for CHWs and how best to linking community based care with primary health care services. There is an urgent need to develop and implement evidence based, cost-effective community based packages that can be delivered on a wide scale. Several community-based newborn care studies have been published from South Asia,5, 6 but as yet none from Africa. In addition, in the context of major investments for HIV/AIDS, an on-going debate of relevance to Southern Africa is how to also maximally benefit maternal, newborn and child survival, linked to these HIV/AIDS investments.

The South African National Department of Health has recognised that in order to increase the likelihood of meeting the health MDGs and to improve all health outcomes, the health system needs to be strengthened through re-engineering of the current Primary Health Care (PHC) system. Lessons learnt from Brazil’s Unified Health Programme/Family Health Programme, led to the revision of the PHC service package in 2010. The PHC approach will change from a passive, curative-oriented approach based on individual health care to a proactive population-based approach utilising community health workers (CHWs).7 Given the national priorities of addressing HIV/AIDS and MNCH, the demonstration of a feasible, integrated service delivery package would provide an opportunity to link to this new national cadre of CHWs and move towards scale.

We conducted a cluster randomized controlled trial in a peri-urban township in South Africa to assess the effect of community health worker visits to pregnant and postnatal women on levels of HIV free survival, and exclusive and appropriate infant feeding at 12 weeks postnatally, and coverage of care and behavioural indicators (antenatal HIV testing, postnatal clinic visit within 7 days of life, uptake of cotrimoxazole amongst HIV exposed infants at 6 weeks, family planning uptake at 6 weeks) and post intervention levels of maternal depressed mood at 12 weeks postnatally.

II. OVERVIEW OF RESEARCH AIMS AND OBJECTIVES

Goal

The goal of this study is to develop, evaluate, and cost an integrated and scaleable home visit package delivered by community health workers targeting pregnant and postnatal women and their newborns to provide essential maternal/newborn care as well as interventions for Prevention of Mother to Child Transmission (PMTCT) of HIV.

Hypothesis

This study aims to test the hypothesis that infants in clusters receiving the intervention, relative to clusters receiving improved health facility care and key information to the mother and family on available social grants and the processes for gaining access to these grants (control clusters) will have significantly improved HIV free survival and higher levels of exclusive appropriate feeding at 12 weeks postnatally.

Objectives

1. To assess post-intervention HIV free survival and levels of exclusive and appropriate infant feeding at 12 weeks postnatally in the intervention and control clusters.

1. To test the hypotheses that the intervention clusters, relative to control clusters, will at the end of the intervention period report significantly better:

1. Uptake of a postnatal clinic visit within 7 days of life

1. Coverage of care and behavioural indicators (antenatal HIV testing, uptake of cotrimoxazole amongst HIV exposed infants at 6 weeks, family planning uptake at 6 weeks)

1. To assess post intervention levels of maternal depression

1. To assess whether the intervention is cost-effective compared to improved health facility care alone. Objective 4 will be evaluated in terms of cost per HIV infection averted and disability-adjusted life years (DALY’s) saved.

SNL Contribution

SNL contributed funds to support data collection and analysis. That mainly included salary for data collectors and managers. MRC, with funding from CDC, supported implementation of the intervention.

III. STUDY SITE

The site for the intervention is Umlazi which is a peri-urban settlement close to Durban in KwaZulu Natal that has a mixture of formal and informal housing. The infant mortality rate is around 60/1000 live births, while there is no reliable figure of the neonatal mortality rate, most estimates place the NMR at about 25/1000 live births. Over 98% of all births in Umlazi (approximate population of 1 million) occur in one major hospital.

The HIV prevalence amongst antenatal clients is 44%. Early perinatal/HIV transmission at 3 weeks was 11.9% in Umlazi. Postnatal transmission was 12.3% in Umlazi, while cumulative overall transmission at 36 weeks in Umlazi was 22.6%. With regard to infant deaths, 25% of the infant deaths (excluding deaths that took place in the hospital or clinic) occurred in the first three weeks, with almost 60% having occurred by six months (Goodstart Study, Unpublished report).

Umlazi is serviced by one main hospital (Prince Mshyeni Memorial Hospital) and a number of primary care clinics. Most clinics are within walking distance of Umlazi residents. Referrals for serious illnesses are usually made to Prince Mishyeni Memorial Hospital.

This site was selected because the Goodstart study team has a long association with the community of Umlazi (over five years). This site was one of three sites that were part of the Goodstart 1 study, which looked at the impact of the national PMTCT programme on HIV free survival at 36 weeks postnatal. In Umlazi the PMTCT programme was shown to have had some impact but there is still room for improvement, notably with regard to appropriate feeding choices.

IV. STUDY DESIGN

Key study outcomes

The study assessed three levels of outcomes:

Mortality, growth and HIV status

1. Neonatal mortality rate (deaths in the first 28 days of life, per 1000 live births)

1. Identification of HIV status in infants at 12 weeks using PCR testing

1. Infant weight and length at 12 weeks

Coverage of care:

1. Co-trimoxazole initiation amongst HIV exposed infants at 6 weeks

1. Clinic visit within the 1st week

1. Having received a postnatal care visit by a CBHW within 48 hours

1. Assessment of biological markers of illness such as diarrhoea

1. Assessment of uptake of HIV specific care and treatment for HIV positive families

Key behaviours

1. Maternal care behaviours of infant cleanliness, warmth

1. Positive behaviours such as the occurrence of immediate (first hour of life) and exclusive breastfeeding at six weeks. Infant feeding patterns will be recorded through 24 hour and previous seven-day recall at 12 weeks.

1. Maternal care-seeking behaviour for infant illness, detection of illness – recognizing signs and severity will be recorded at 12 weeks

1. Use of dual protection methods for family planning and HIV /STD prevention recorded at 12 weeks

1. Maternal mental state assessed at 12 weeks

In addition, key process documentation will include an assessment of the process involved in training and supervising community workers, and the retention of CHW’s and factors affecting motivation and intensity of the intervention (coverage of scheduled visits).

Study design and methods

This study used a cluster randomized unblinded controlled trial design. The trial consisted of 30 randomized clusters (15 in each arm) that each contained a sufficient number of women delivering babies in order to obtain an adequate sample size to determine differences between the intervention and control sites.

The GIS unit at MRC divided Umlazi into clusters based on renumeration areas from the 2001 census. Areas of Umlazi had to excluded because they were too dangerous or too sparsely populated. Following the drawing of clusters, a baseline survey (or pre-randomization survey) was conducted to check the homogeneity of clusters and determine if clusters would need to be stratified. The baseline was a household listing and household census in order to map the area and establish numbers of women of childbearing age in each household. The household listing was conducted according to Statistics South Africa (STATSSA) Master Listing Guidelines (STATSSA, Master Sample 2004 Listers Manual, Pretoria). The household census was a simple listing of the gender and age of all household members, and selected socio-demographic and mortality data. Results of this exercise established the homogeneity of clusters and neither stratification nor matching was done.

Thereafter, pregnancy registration was conducted in all clusters quarterly in order to ensure that all eligible women in a cluster were identified. This was done by CHWs in the control and intervention clusters. Satellite maps were created for each cluster to help the CHWs find all housing in the area.

Sample size calculation

To increase HIV-free survival from 74% to 84%, with 80% power then an individually randomised trial would require 279 live births per arm. Assuming an ICC of 0.04 for a cluster randomised trial study required 750 HIV exposed children per arm (50 per cluster) with a design effect of 3. Loss to follow up of approximately 20% was added to this sample size. Based on an HIV prevalence rate amongst pregnant women of 40%, sample size was calculated to be 120 pregnant women per cluster.

IRB approval process

The study received ethical approval from the Medical Research Council Ethics Committee on 5 May, 2008.

Study management

This study is managed by a team of investigators at the Medical Research Council, University of the Western Cape, University of Stellenbosch, and Save the Children.

In the study protocol, a DSMB was planned to meet halfway through the completion of the study to review study safety. However, a DSMB was not convened due to funding cuts. The PI met with the MRC Ethics Committee to discuss what the study team would be required to submit to explain this protocol deviation. There were no recommendations made to establish or convene a DSMB given the low risk nature of the Goodstart intervention.

Key implementation dates

See timeline in Appendix 2.

Cohort

The study cohort included all pregnant women and their newborns identified in the study clusters during the recruitment period (June 2008-December 1, 2010) and gave informed consent for study participation.

Questionnaires

The study endpoint was 12 weeks after birth. Data collectors interviewed mothers using a questionnaire that included questions on socioeconomic status, antenatal care, intrapartum care, postnatal care, HIV, infant feeding, illness and care seeking, mental health of the mother, vaccination, and child grants. Data collectors used other questionnaires in the event of miscarriage/stillbirth, neonatal/infant death or when it was necessary to interview a caregiver other than the mother. These questionnaires collected a subset of the information collected at 12 weeks, and some additional information relevant to the situation. Data collectors also completed an enrolment form on the day of enrolment. In addition, data collectors regularly visited the hospital where most deliveries occurred to ensure study staff knew about births as early as possible, and completed a birth capture form.

Data collectors were trained to collect dry blood spots (DBS). DBS was collected from study infants of HIV infected mothers by means of a heel prick during home visits at 12 weeks and tested using DNA PCR testing.

Community health workers completed questionnaires specific to each visit or a missed visit questionnaire (separate questionnaires for the intervention arm and control arm). Questionnaires included questions on the health and behaviours of mothers and babies before or after birth.

Questionnaires were completed using mobile phone technology, which is described below. Questionnaires were programmed into mobile phones as well as logical, range, missing data and other checks to minimise data entry error by field workers.

Data entry

Data collection for this study used new mobile phone technology. Both Community Health Workers delivering the intervention package and Data Collectors were trained to use this technology and to directly enter data using their mobile phones. Therefore, there is no data entry required for this study. The Data Quality Manager is responsible for reviewing data for errors and data cleaning.

The use of the mobile phone technology has been described in an article published by the study team in 2009 in BMC Medical Informatics and Decision Making (http://www.biomedcentral.com/1472-6947/9/51).

Implementation of the mobile system ("Mobile Researcher") was a partnership between the Health Systems Research Unit at the Medical Research Council of South Africa and a private digital solutions company, Clyral. The only hardware requirement for Mobile Researcher is that the mobile phone must be enabled for the Java programming language. The software consists of a Java 2 Micro Edition (J2ME) client which is installed on participating project staff handsets and communicates with the Mobile Researcher web application, also developed on the ASP .NET 3.5 framework. Java's core code is available under open-source distribution terms. We established a web-based system that allowed electronic surveys or questionnaires to be designed on a word processor, sent wirelessly to standard entry level mobile phones, and then used in interviews. Survey completion takes place offline, and no network coverage is necessary.

Completed surveys are automatically uploaded to the host computer. If there is no mobile network coverage, completed surveys are stored securely until a signal is found at which time completed surveys are uploaded. Entry level mobile phones can store approximately 50 completed standard-length surveys. Mobile Researcher can incorporate multiple choice, free text, numeric, date, time and other question types.

In addition, Mobile Researcher can also accommodate branching and skip logic as well as enforced validation in the field. Survey data are uploaded using low cost general packet radio service (GPRS). A web-based interface was developed to facilitate the review and exporting of results in standard file formats such as comma separated values (CSV) and Microsoft Excel. Figure 3 is an illustration of an Excel report that provides formatted information files on data collector, time taken for the participant id and household code and when the actual upload of the data took place. Outputs such as this (Figure 3) can be generated as often as the investigator would like. Alternatively, built-in graphs and reports on the web-based interface permit real-time visualization of survey responses. Supervisors can communicate with data collectors directly, either through a call to the mobile phone or through SMS messaging. For logistic and supervisory purposes SMS messages can be sent through the web-interface to one or more data collectors at the same time.

All survey data were encrypted, thus maintaining the confidentiality of responses. Communication between the browser and the server was encrypted using 128-bit SSL. System servers were secured by firewalls to prevent unauthorised access and denial of service attacks, while data was protected from virus threats using NOD32 anti-virus technology. Access to the web-interface is protected by passwords. In the current study, access to the data was restricted to the principal investigator, the project manager, data quality officer and web administrator.

Data analysis

For the primary outcome (HIV free survival), an intention to treat analysis was performed. A generalised linear model was used to estimate the risk ratio of the intervention relative to the control arm as the statistical measure of the trial. Confidence intervals (95%) were also estimated and they took into account the cluster design of the trial. Analyses can be found in the main outcome paper (forthcoming).

V. OVERVIEW OF INTERVENTION PACKAGE

The intervention was a structured programme of home antenatal and postnatal visits with specific content covered at each visit. The home visits were done by trained Community Health Workers (CHWs). In each intervention cluster, there was one CHW who covered all households in the cluster. CHWs identified pregnant women living in their intervention area and visited the women at home according to the intervention schedule. The content and conduct of the intervention is specified in a training manual, linking to existing PMTCT, IMCI and newborn care guidelines.

Minimum qualifications for CHWs included: women age 20-60 years, completion of grade 12 (passed Matric), literate in English and Zulu, residence in Umlazi for at least 2 years, and experience in working in health in the communities. Additional qualifications can be found in the Manual of Operations. Job descriptions were advertised and CHWs were interviewed and hired by study staff. Community leaders gave recommendations on qualified candidates.

Number and timing of pregnancy and PNC visits

Intervention arm:

· Recruitment and enrolment at 24 weeks or later in pregnancy (Women identified in earlier stages of pregnancy were supposed to be enrolled at a later date, though some were enrolled at that time. Women enrolled late in pregnancy making it difficult to complete pregnancy visits)

· 2 antenatal visits

· AN1: 14-28 weeks

· AN2: 29-39 weeks

· 5 postnatal visits for normal weight babies

· PN1: 24-48 hours after discharge (initially, used time since birth, but that

· PN2: day 3-4

· PN3: day 10-14

· PN4: 3-4 weeks

· PN5: 7-8 weeks

· 2 extra PNC visits for low birth weight babies:

· Extra visit 1: During the first week back at home

· Extra visit 2: During the first week back at home

Initially, the timing of visit was counted from the day of birth. But it was too difficult to visit women and newborns 24-48 hours after birth because many women stay in the facility for approximately 24 hours after birth. Therefore, the visit time was measured from day of discharge.

Control arm:

· Recruitment at 20 weeks (women need to engage early in order to receive the child support grants)

· 1 pregnancy visit: before 24 weeks

· 2 PNC visits

Content of pregnancy and PNC visits in the intervention arm

Antenatal visit 1 at home

1. Emphasise the importance of at least 4 antenatal clinic visits

1. Discuss illnesses pregnant women can get

1. Nutrition

1. Avoid smoking and alcohol

1. Taking all supplements hospital provides

1. Importance of HIV test

1. Birth plan

Antenatal visit 2 at home

1. Importance of HIV test

1. Discussion of medicines that need to be taken if woman is HIV+ (women are not required to disclose HIV status to CHWs)

1. Danger signs for pregnant women

1. Birth plans

1. Discussion of feeding options

1. Breastfeeding counselling, including the importance of breastfeeding, information about exclusive and immediate breastfeeding, attachment

1. Why babies cry

1. Importance of Road to Health card

Postnatal visit 1 at home (24-48hrs)

1. Discuss illnesses postpartum women and newborns can contract

1. Feeding frequency

1. Hygienic cord care and eye care

1. Thermal care and Kangaroo care if needed for preterm babies

1. Observe breastfeeding, help with attachment and how to prevent sore and cracked nipples

1. Input on exclusive breastfeeding and stress importance of breastfeeding

1. Discuss danger signs

1. Discussion of medicines that need to be taken if woman is HIV+ (women are not required to disclose HIV status to CHWs)

1. Input on knowing your baby

1. Stress importance of going to clinic 6 days after birth

Postnatal visit 2 at home (3-4 days)

1. Steps from PN1

1. Input on handwashing

1. Input on how to respond to advice on supplemental feeding

1. Importance of testing baby for HIV at 6 weeks using PCR (earlier if child is sick) and cotrimoxazole from 4-6 weeks if mother is HIV+ (Women are not required to disclose HIV status to CHWs)

1. Discuss postnatal blues

Postnatal visit 3 at home (10-14 days)

1. Steps from PN1 & PN2

1. Continue to discourage solid foods

1. Importance of playing with baby

1. Input on planning follow up visits at clinic to check baby’s health

Postnatal visit 4 at home (3-4 weeks)

1. Steps from PN1 – PN3

1. Formula sustainability

1. Clinic visit at 6 weeks for family planning and child immunization

1. Counsel on feeding options after baby is tested for HIV:

3. If baby is found HIV+: continue breastfeeding.

3. If baby is found HIV-: support mother to implement feeding option that most suits her circumstance

1. If mother HIV- or unknown HIV status: ECF for 6 months, continue BF for at least 2 years.

1. Discourage mixed feeding

Postnatal visit 5 at home (7-48weeks)

1. Steps from PN1 – PN4

1. Look at infant weight from 6 week visit and discuss implications

1. Family planning input

1. If mother HIV+, make sure baby has been tested for HIV at 6 weeks using PCR and counsel on feeding.

Extra visits for low birth weight infants (7-8 days and 16-19 after birth)

1. Steps from other PN visits

1. Discuss thermal care and Kangaroo care (KMC)

1. If using KMC, make sure baby is positioned correctly

1. Input on how small babies need frequent feeding but are sometimes too tired to cry to tell you they are hungry

1. Feeding frequency

Content of pregnancy and PNC visits in the control arm

In control clusters, CHWs provided key information to the mothers and families on available social grants and the processes for gaining access to these grants.

Referral system

One of the roles of CHWs was to counsel mothers on early identification of newborn danger signs and referral to health care facilities where necessary to prevent complications. If danger signs were detected, a CHW wrote a referral note to the local primary health care (PHC) clinic. A follow-up study on babies referred by CHWs was conducted during implementation and found that 6% of babies where referred (148/2423) and referral compliance was 95% among those that agreed to be interviewed during the follow-up study (104/110). Difficulty breathing and rash accounted for the highest number of referrals (26% and 19% respectively). More detailed results will be available in a forthcoming paper (manuscript submitted as of May 2012).

Supervisory structures

The CHW supervisory structure was changed several times during study implementation. In 2010, the structure became one supervisor and one assistant supervisor for CHWs in the control arm and the same structure for the intervention arm. The supervisors provided supportive supervision to the CHWs through weekly planning and monitoring meetings. These meetings also addressed the challenges encountered in implementing the project and provided CHWs with the opportunity to share experiences.

During scheduled shadow visits, supervisors observed CHWs during home visits and afterwards provided mentoring and coaching on how to improve and positive feedback. Supervisors also did spot visits to confirm CHW visits. Supervisors completed spot visit and shadow visit forms using the mobile phone technology. Supervisors also had forms for assessing the work of the assistant supervisors.

The data collector manager, quality manager, and project manager oversaw the 4 data collectors and held weekly meetings and oversaw ongoing training.

The mobile phone system was part of the supervisory structure. Supervisors were able to access a report for each CHW that showed overdue visits. Supervisors investigated overdue visits to see if there is a reason for the late visit, such as woman has not yet delivered. Supervisors were also able to view the participant’s summary, which showed the number of completed visits and attempted visits to see if the CHW reported an attempt to visit the participant home but was unsuccessful in interviewing the participant. To verify attempts, supervisors called participants.

The system supported appropriate management by showing the number of visits that CHW has scheduled in the coming week based on the date of expected delivery (if the birth has not yet been captured) or the date of discharge (if the actual date of birth has been captured). The CHWs planned their schedule based on this output and management could adjust the workload for each CHWs.

Quality assurance mechanisms

We employed a full time quality control person (Ms Yages Singh) for the data collected. At various times during the course of the project she was aided by assistants in order to ensure data quality. Ms Singh also had regular meetings and training throughout the course of the project with data collectors as well as CHW’s in order to ensure quality data.

To ensure the quality of the intervention we developed a comprehensive management and supervision system that combined mobile phones with a web-based interface. The system was designed to integrate all CHW activities into a web-based narrative that is initiated the moment a participant is recruited with all intervention and data milestones triggered automatically. Computerized algorithms were designed to monitor and trigger the intervention and assessment protocols to aid in the real-time supervision and management of CHWs in terms of intervention fidelity, quality control, visit planning and the monitoring of CHW productivity. The web based management system was developed in partnership with a private digital company in South Africa, Clyral. All the CHWs were women with no previous administrative or data collection experience, but all had previous experience using basic mobile phone functions such as calls and SMS. Training CHWs in the use of the mobile phone application consisted of a one day training workshop, followed by several one hour refresher courses throughout the duration of the project. There was also weekly in-person group supervision, with individual supervision when necessary. Each CHW was supplied with her own project mobile phone.

The underlying motivation was to develop a system of daily monitoring of CHW visits that to aid in the supervision of a large cadre of CHWs. Specific focus for the system were:

1. Recruitment of study mothers, CHW assessment, and adaptive scheduling and visit planning;

2. Intervention delivery support, fidelity monitoring and individual case monitoring;

3. Caseload reporting and management

1. Recruitment, assessment, and adaptive scheduling and planning

CHWs recruited pregnant participants in clusters by going door to door. The GPS captures the geographical coordinates on the CHWs mobile phone for each household visited. These data facilitate monitoring of whether CHWs are visiting all households in their designated areas. As a CHW enters a household, the CHW begins monitoring the visit duration with one click on the phone. Then, as she leaves the home at the end of the intervention visit, she repeats the click on the phone. GPS- and time- stamping functions are built into each mobile phone, allowing location and duration to be assessed and transferred in real-time to the web-based supervision and management system. The GPS function reduces the possibility of falsifying data substantially, as the time between visits and the location of the visit is automatically recorded.

For each visit that a participating mother is due to receive, a dated milestone is created on her participant page in the web-based management Console. The dated visit milestones for all mothers in the program filter through to a weekly timetable for each CHW. This timetable details which mothers she needs to visit that week. At each visit the CHW records the duration and content of the visit on her mobile phone, which ticks off each milestone in the participant’s lifespan as completed. Once completed, the visit reminder is removed from the weekly timetable. As soon as the woman gives birth, the birth date and time is entered into the mobile phone and transferred to the web-based system, which automatically schedules and triggers on the phone all postnatal visits. Simultaneously, the CHW and the CHW supervisor will receive a SMS notification of the birth and the date of all scheduled visits. The GPS coordinates of each household could also be used in such a way that within each week, those mothers who live closest to each other are scheduled for appointments on the same day.

2. Intervention delivery support, fidelity monitoring and individual case monitoring

We developed a series of short questionnaires that were triggered on the mobile phones when the CHW entered the identification number of the household being visited. This questionnaire serves as a reminder about the core intervention messages to be covered at that visit, a check on intervention fidelity that the topics were covered, and as the data source for the adaptive scheduling of follow up visits and intervention content. Antenatal questions include whether the woman has made a booking at the antenatal clinic; her feeding plan is and whether she has been tested for HIV. Postnatal questions include questions about maternal and neonatal illness; whether the baby received AZT; whether the mother went to the clinic at 6 days post-birth; breast health; whether the infant has been tested for HIV; and if the baby is HIV+ whether he or she is receiving cotrimoxazole.

In order to account for day to day participant availability, the system allows a CHW to suspend a visit based on the availability of the participant. If the CHW determines that she should return to complete a visit at some other time in the future, she can tick the appropriate box on the mobile phone. Her tick is immediately registered on the web based system. This is also the case for missed visits (participant not home when CHW arrives) or visits cut short for other reasons (e.g. irritable infant). In this way, the system is able to monitor all aspects of the intervention such as CHW caseloads, clusters with higher levels of missed visits as compared to other clusters; reasons for missed or suspended visits; and CHW performance in real time on a daily basis. The web-based management and supervision system (and mobile phone connectivity) allows supervisors to (remotely) access the Management Console and monitor real time information about each participating mother and child in the intervention program, as well as information on each CHW’s activities and caseloads.

The Management Console has a live report of all visits which have taken place, which can be filtered by date and by CHW, in order to establish how many visits are happening on any given day or week by any given CHW. The Console also generates a ‘visit-time’ report , which details for each participant in the intervention, not only how many visits they have had, but also the cumulative amount of time for all visits combined. In this way, one can easily identify outliers among participants or among CHWs. Based on these reports, together with the system flags that are generated, supervisors are able, in real time, to schedule a ‘shadow’ or ‘spot’ visit’ for the supervisor to conduct checks on the quality of home visits. In addition, supervisory visits are randomly generated and scheduled by the Management Console based on each weekly schedule for each CHW. These visits allow the supervisor to validate the observational reports of the CHW or identify households which should receive extra support.

The management system can be used to optimize supervision at every level of the intervention program including:

· Recruitment and scheduling of visits

· Monitoring activity and progress of CHWs

· Managing health information and emergencies

· Accessing data such as date of visit to household and duration of visit

· Provide knowledge of visit schedule

· Pinpoint discrepancies and anomalies automatically with alerts sent to supervisors

3. Caseload Reporting and Management

The Management Console has a live report that allows supervisors/managers access at any time to the current caseload of each CHW working in the project. This aids in the planning of staffing needs, as the system is able to send alerts when a CHW’s caseload approaches a maximum threshold. The caseload report is able to differentiate between active and inactive cases for each CHW, allowing the report to give an accurate picture of each CHW’s workload. When a participating family has temporarily left the area or declines to receive visits for any given period of time, their status in the system can be changed from active to inactive. The caseload reports detail only those families who are actively being visited by the CHW. Detailed information about the nature of each caseload is also summarized in the report. It is also possible to view the caseload filtered by any type of risk factor of interest, which is helpful in determining workloads when families or mothers meeting certain criteria qualify for extra intervention visits. The system also provides the data to aid the ‘face to face’ supervision that takes place about particular ‘cases’ and ‘problem areas’.

Innovations

We engaged women using motivational interviewing by assisting them to engage with social supports, at the sensory level (behaviour change materials and media) and at the cognitive level through problem solving, three of the four levels recently shown to be most effective in engaging participants in ‘best practice’ behaviour change programmes. 8 This EBF rate (almost 30%) is 3 times the rate of the two most recent DHS surveys in South Africa as well as the rate of EBF achieved in the PROMISE-EBF trial using single purpose CHW’s focused solely on breastfeeding counselling (Figure 3) 9-11. The fact that a similar doubling of EBF (albeit from a higher base rate) was achieved in the context of an integrated package of care, rather than one focussing only on feeding counselling is important for promoting the efficient and effective use of human resources in the South African health system.

Role of MOH and other partners/stakeholders

Throughout the course of the project there was an ongoing dialogue between the project and Prince Mshiyeni Memorial Hospital authorities and provincial health authorities. We attended a number of provincial stakeholders meetings during the course of the project. Discussions at these meetings included different strategies for HIV prevention and treatment were discussed including phasing out the free formula milk, provision of ARVs to HIV pregnant women and their newborns, and male circumcision.

Over the course of the project we held over 50 meetings with different stakeholders, including ward councillors were held as part of the process to mobilize and lobby the Umlazi community to support Good Start. We established a Community Advisory Group (CAG) which was elected by the community and ward councillors. The CAG worked closely with Goodstart as well as other community initiatives such as a group of GOGO’S (grandmothers), a group of Mkhulu’s (grandfathers) together with a men’s NGO called Brothers for Life. The CAG met on average five times per year.

VI. STEPS TO IMPLEMENT RESEARCH STUDY

Formative research

For the Goodstart 3 study we did not conduct a formal formative study. Two activities however can be considered formative in nature:

1. Goodstart 3 followed the implementation of Goodstart 2/PROMISE study from which important formative lessons were learnt. The most important was the decision to use motivational interviewing in order to more directly confront community and household barriers to the implementation of exclusive and appropriate feeding. Anecdotal evidence from Goodstart 2 was that many women that had been counselled were determined to exclusively breastfeed, but were confronted with barriers such as pressure from partners, mothers and mother-in-laws to mix feed and felt unable to implement their choice. In designing the Goodstart 3 intervention we felt that it was essential that the intervention did not stop at simple feeding counselling, but that the intervention dealt with these barriers as well.

2. We conducted a baseline survey of all clusters prior to cluster randomisation in order to check the homogeneity of clusters. This involved a household listing and household census of the 30 clusters in order to map the area and establish numbers of women of childbearing age in each household. The household listing was conducted according to Statistics South Africa (STATSSA) Master Listing Guidelines (STATSSA, Master Sample 2004 Listers Manual, Pretoria). The household census was a simple listing of the gender and age of all household members, and selected socio-demographic and mortality data.

Material development and testing

Comprehensive training manuals for trainers and trainees were developed by the study technical staff based on materials from UNICEF and WHO and local experience. The manuals went through a rigorous revision process to ensure that the final products covered all the relevant topics the CHWs needed to know in order to carry out the study effectively. In addition, counselling booklets for CHWs in the intervention and control arms were developed in order to assist CHWs through each home visit. The booklets have general guidance and the key messages that the CHW needed to communicate at each visit. Study materials were developed through a consultation with community members, CHW’s, professionals involved in Goodstart 2 and design specialists at the Medical Research Council.

Pilot testing process

The manuals were not pilot tested prior to study implementation.

Future of these materials

The South African National Department of Health has recognised that in order to increase the likelihood of meeting the health MDGs and to improve health outcomes, the health system needs to be strengthened through re-engineering of the current Primary Health Care (PHC) system. Lessons learnt from Brazil’s Unified Health Programme/Family Health Programme, led to the revision of the PHC service package in 2010. 12 The PHC approach aims to shift from a passive, curative-oriented approach based on individual health care to a proactive population-based approach utilising community health workers (CHWs).7 The Goodstart 3 intervention manual developed for the study has been used as the foundation of the Primary Health Care system re-engineering that is currently being rolled out. Dr Tanya Doherty one of the Goodstart co-principal investigators is consulting with the National Department of Health on the further development of materials for the national roll out.

Health facility strengthening and Community mobilization

During implementation, CHWs reported that mothers referred to clinics for a 6 day visit per study protocol and government policy were turned away by the clinics or received an inadequate check and were told to return for the 6 week immunization visit. Other mothers told CHWs they would not go to clinics at 6 days, knowing they would be turned away at that time. In addition, there was apparent mistrust of CHWs both from clinic staff and communities. In the most extreme cases, CHWs were held up at gun point and two project cars were hijacked and there were other indications of general suspicion of study staff within the community and local political leaders. Leadership at the District Department of Health questioned the qualifications of the CHWs, and clinic staff scoffed at referrals from CHWs.

To address these problems, the study staff used Save the Children’s Partnership Defined Quality (PDQ) methodology in 2 clinics, to bring health staff and communities together to identify and improve quality in care at facilities in the study area. This tested methodology requires separate discussions with community groups, including clinic users and no-users as well as underrepresented groups and influential members, and clinic staff (Exploring Quality). Then representatives from community and clinic groups are brought together to discuss their definitions of quality and the problems they identified that prevent them from reaching quality (Bridging the Gap). This stage may also include social activities for clinic staff and community members to feel more comfortable with each other. Finally, quality improvement teams are formed with both clinic and community members to identify solutions to those problems (Improving quality). (http://www.coregroup.org/storage/documents/Diffusion_of_Innovation/PDQ-Manual-Updated-Nigeria.pdf)

See attached PDQ report for full description.

Trainings

CHWs for the control and intervention areas were trained separately. Training for CHWs in the intervention area was 10 days while training for CHWs in the control area was 2 days. CHWs in the intervention arm were trained on counselling and communication skills, antenatal care, baby feeding, mother-baby interactions, postnatal depression, neonatal care, counselling on disclosure. CHWs working in the control area were trained to assist households to understand the processes involved n accessing government social grants, in particular the child support grant.

A pre and post-test was used to measure the effectiveness of the training and the knowledge that the CHWs obtained through the trainings. Immediately prior to starting the intervention, CHWs in the intervention area underwent a 3 day refresher course and CHWs in the control area underwent a 1 day refresher course.

Trainers were senior Health Systems Research Unit staff with experience in CHW training, breastfeeding promotion, PMTCT, neonatal care, training communities in accessing child grants and community development. Training materials have been shared with the department of health.

Monitoring and Evaluation

Study monitoring

The computer system used for data collection provided output on the number of completed visits and visit type, scheduled visit dates (antenatal visits scheduled based on expected date of delivery, postnatal visit dates are based on discharge date), actual visit dates, expected date of delivery, actual date of delivery, and date of discharge. Because information collected on cell phones is immediately transmitted, data is available in real time. Scheduled visit dates were based on the expected date of delivery; once birth capture data has been collected and the actual date of delivery is known, the scheduled dates for postnatal visits are updated by the computer system.

This information allows study supervisors and management to monitor overall study progress towards recruitment targets as well as daily outputs of each CHW and data collector. The system is designed to flag inconsistent data, such as inconsistent visit type/PID or duplicate visits, in real time. The computer system allowed study management to look at the data in different views: by cluster, by individual CHW, and by participant. These different views allow managers to see any activity that is unexpected or outside the norm.

GPS data was recorded for each participant to facilitate tracking of participants. Study field assistants collected GPS information on the location of participants home after recruitment so that the locations of participants’ homes could be mapped to see the distribution of participants within the cluster.

The Data Collector Supervisor received weekly reports on the expected deliveries and the number of scheduled 12 week visits based on date of delivery. If the expected deliveries were not captured within 2 weeks after it appears on the Data Collector Supervisors’ report, the participant was called to find out if she has delivered outside of Prince M hospital. This system was intended to reduce the number of missed early postnatal visits.

Evaluation activities

· Interviews with mothers, or other caregivers, 12 weeks after birth (earlier in the case of miscarriage, stillbirth, or death)

· Birth capture data from hospital records

· HIV testing of HIV+ mothers and their infants using dry blood spots and PCR testing

· CHW Knowledge Assessment: CHWs received a 10 days training in the intervention area with pre and post-tests. CHWs also had ongoing refresher courses followed by post-test assessments.

VII. COSTING

Cost data were collected using the Cost of Integrated Newborn (COIN) Care tool, a standardised, Excel-based tool designed to collect the added cost of facility and community based newborn care in the context of MNCH programmes. The tool was developed by MRC South Africa with Saving Newborn Lives, Save the Children and was used across several studies in Africa, Asia and Latin America. The COIN care tool was used to record cost data for start up and running costs for one year. Analysis is ongoing and results will be included in the main outcome paper as well as a separate cost analysis paper.

VIII. STUDY MANAGEMENT

Overall responsibility for the study was that of Professor Tomlinson. At the start of the study this was shared with Professor Mickey Chopra. Professor Chopra left the Medical Research Council in 2009 to take up a position of Chief of Health at UNICEF New York. Following this departure Professor Tomlinson established an informal management structure that included Dr Tanya Doherty and Professor Debra Jackson. This was especially important when Professor Tomlinson took up a position at Stellenbosch University in 2009. Administrative issues at the Medical Research Council were then handled by Dr Doherty.

At the start of the study there were two senior managers at the Umlazi site – Ms Petrida Ijumba and Ms Thoko Ndaba. During the course of 2009 the two of them experienced significant conflict with one another to the point that communication between the two ceased. The conflict became so serious that an external management consultant was hired in order to try to resolve the issue. This was not successful. As a result Ms Ndaba was removed from her position as supervisor and manager of the CHW’s and took over the health systems strengthening component of the study.

Perhaps the main decision that had to be made about the study occurred in mid-2010. In July 2010, without warning, the Centers for Disease Control stopped all grants to the Medical Research Council until further notice. This was not directed only at the Goodstart 3 project but to all projects at the MRC funded by CDC. The net result of this was that as of July 1 2010 all funds for the Goodstart 3 project from CDC dried up. A decision was made at this time to stop further recruitment. Professor Tomlinson, Professor Jackson and Dr Lawn conducted a new sample size calculation and it was determined that the sample size at that point was sufficient for the primary outcomes. Follow up of women and infants already recruited occurred because of discretionary funding from the Director of the Health Systems Research Unit, Medical Research Council.

IX. KEY FINDINGS

There was no statistically significant impact on HIV free survival, the main outcome. There were statistically significant changes in exclusive breastfeeding at 12 weeks, clinic visit within first week, birth preparedness, and knowledge of newborn danger signs. EBF was particularly important given that increases were greater than changes seen during the Promise EBF (aka Goodstart 2) study that focused primarily on breastfeeding. About 40% received all 7 visits about 40% received a visit within 48 hours of discharge.

SNL key indicators are included in tables at the end of the report. These key findings will be included in the forthcoming outcomes paper (to be submitted to Lancet in early July)

· Challenges and lessons learned from implementation

X. DISSEMINATION

Dissemination activities have not been conducted. An event at Prince M hospital for local stakeholders is planned for later in 2012. SNL will support a policy brief after the outcomes paper has been submitted. Should the main outcomes paper be accepted into Lancet, a national launch event will be held.

Published papers

Two papers have been published on the protocol and mobile phone use for data collection

Mobile phone:

http://www.biomedcentral.com/1472-6947/9/51

Protocol paper:

http://www.trialsjournal.com/content/12/1/236

Planned or submitted papers

See paper queue in annex 3.

XI. REPLICATION

While community workers in this study were part of a research study rather than government hired and trained generalist community workers, the training, implementation and supervision of the community workers approximated the intervention planned by the South African government in their expanded primary health care reengineering. In addition, the package promoted linkages between community level and existing health systems. This CHW programme is currently being implemented by the South African government in close consultation with a number of Goodstart investigators (Professor Jackson, Dr Doherty, and Professor Tomlinson). The Goodstart intervention has informed all levels of this process, in addition to establishing the effectiveness of such an integrated package. Implementing an effective scaled up and integrated package for newborn care, PMTCT and infant development using generalist health care workers in urban informal settings has significant implications for country wide integration of programmes. In addition a rigorously designed and well disseminated study would have implications beyond South Africa, particularly for linkages with PMTCT and HIV/AIDS programmes. Furthermore, there are few well described models of successful approaches to reach the urban poor with MNCH care.

XII. CHALLENGES AND NEXT STEPS

Challenges

Working in urban areas

Working in an urban area presents unique implementation challenges. The population is highly mobile and participants often move residences to stay with different family members or partners. This mobility not only took time and money to track participants, it also made it more difficult for CHWs to build relationships with other members of the household who could support the woman. Participants who were employed or still in school were often not at home so CHWs had to return for follow-up visits. Violence was extremely problematic. During the study period, two CHW’s and one data collector died, 3 motor vehicles were hijacked at gunpoint, 11 mobile phones were stolen and 8 CHW’s were mugged while performing work duties. We believe that this is the reality of any scale-up in the South African context and are the very challenges that will be faced by the re-engineered primary health care system.

Technology and supervision

The mobile phone technology was new and under development at the start of the study for both the providing company, Clyral, and the project staff. At the start of the study, staff did not know how much information they needed to monitor the study and wound up with more information than necessary. Changes had to be made along with way with constant dialogue between Clyral and project staff.

The supervisors and management receive daily emails summarizing the data collection that has occurred. Some staff were uncomfortable with email and did not check these reports daily and had difficulty following up due to the number of notices that appeared in the system. The technology made is easier for late visits to be immediately visible for all supervisors and management. While it demanded more accountability, it also created tensions within the staff and showed weaknesses in the supervision.

Policy changes

ART policy:

Policy has changed and women are now started on ART earlier during pregnancy (at 14 weeks). This change will impact uptake of cotrimoxizole, a study outcome. In addition, it changed the counselling that CHWs had to give to participants, causing some confusion and re-training.

Formula feeding policy:

The South African government provided free formula to HIV+ mothers to allow mothers to choose to exclusively feed based on their needs and preferences. When the policy was not shown to be effective in reducing mother to child transmission due to imperfect implementation of exclusive formula feeding, the provincial government of KZN (Department of Health) decided in 2010 to phase out the free provision of formula (based on WHO recommendations released on World AIDS Day in 2009). At the start of the study, CHWs counselled participants on exclusive feeding and informed them that they could obtain free formula from clinics. Since the policy change, CHWs have changed their counselling to reflect the provincial policy. However, participants report that they still get formula form the clinics and it is evident that the district has not stopped distributing formula. The study staff met with provincial officials, who were surprised to learn that the district clinics continue to distribute formula. Still, clinics continue to distribute formula, making it difficult for CHWs to appropriately counsel participants.

Health facility challenges

Per study protocol, CHWs refer participants to clinics for a postnatal visit one week after birth, and uptake of a 6 day visit is a study outcome. However, CHWs reported that participants were turned away from the clinics. Based on anecdotal reports, some participants delayed seeking care at clinics for sick newborns because they feared disclosure at the clinics, feared abuse from nurses, did not want to face long waits or stock-outs of drugs, or sometimes attempted to seek care but found the clinic was closed early. These health facility weaknesses led to the decision to use Save the Children’s PDQ methodology, as described above.

Disclosure

Participants may find it difficult to disclose to family members and partners, many of whom are not husbands or stable partners. The study encourages family participation in CHW visits, especially grandmothers who are involved in the care of the newborn. CHWs report that some families do not participate. Failure to disclose may explain why participants do not want their families to know about their participation or do not want family to hear the counselling sessions. CHWs have also faced difficulties contacting members of the family when trying to track participants for missed visits. Information on disclosure is collected at the 12 week visit and failure to disclose can be compared to failure to adhere to PMTCT and breastfeeding choice. Previous research has shown that failure to disclose means women may not be able to respond to cultural pressures and have trouble adhering to exclusive feeding.[footnoteRef:1] Women that do not disclose have little or no support for their choices in caring for their newborns. [1: Doherty T, Chipra M, Nkonki L, Jackson D, Greiner T. Effect of the HIV epidemic on infant feeding in South Africa: “When they see me coming with the tins they laugh at me.” Bull World Health Organ 2006; 84: 90-96..]

Disclosure to CHWs is not documented, but CHWs report that many women do not disclose their status. Failure to disclose to CHWs made it difficult for CHWs to appropriately counsel women on breastfeeding and PMTCT.

Human Resources challenges

During the study period, two CHWs and one data collector died, 3 motor vehicles were hijacked at gunpoint, 11 mobile phones were stolen and 8 CHWs were mugged while performing work duties. We believe that this is the reality of any scale-up in the South African context and are the very same challenges that will be faced by the re-engineered primary health care system. As mentioned above we also experienced significant human resource challenges between the two senior Umlazi site managers.

XIII. APPENDICES

1. Results (SNL global indicators)

2. Study Timeline

3. Paper queue

XIV. ATTACHMENTS

1. Published articles

2. PDQ report

3. IRB approval letter

4. Protocol

5. Tool kit

· Manual of Operations (includes SOPs)

· Training materials

· CHW Counseling booklets

· Questionnaires

21

September 2012

Appendix 1: RESULTS (SNL GLOBAL INDICATORS)

Intervention area

Control area

Total

Enrollment:

Screened

Enrolled

Enrollment rate1 (%)

Retention:

Women interviewed 12 weeks after live birth

Woman interviewed after infant death

Woman interviewed after miscarriage or stillbirth

Caregiver interviewed

Total number of interviews

Retention rate2 (%)

1894

1821

96.15%

1659

28

49

2

1708

94%

2243

2136

95.23%

1902

34

63

3

1965

92%

4137

3957

95.65%

3561

62

112

5

3673

93%

*Data from recruitment, enrolment, 12 week questionnaire, neonatal/infant death questionnaire, miscarriage/stillbirth questionnaire, and caregiver questionnaire.

1Enrollment rate = Enrolled / Screened

2 The final data collection point was 12 weeks after birth. However, women were interviewed earlier if they experienced a miscarriage/stillbirth or the infant died before 12 weeks of age. A caregiver may have been interviewed if the mother could not be interviewed. The study had separate forms for these scenarios. Retention rate = Total # interviews / Enrolled.

BACKGROUND CHARACTERISTICS

Table 2.1: Background characteristics

Percent distribution of interviewed women by selected background characteristics

(N=women interviewed at 12 weeks) N= 3673

Intervention area =1708

Control area =1965

p-value

Age (years)

<20

20-24

25-29

30-34

>35

Don’t know/Missing

Mothers age median

Education

None

Some primary (1-6 grade)

Some high school (7-12 grade)

Tertiary


Marital status

Single/divorced/widowed

Married

Cohabiting


Parity

Primipara

Multipara


n

386

680

354

197

91

0

23

6

62

1534

104

2

1494

78

134

2

834

872

2

N

842

1414

816

407

194

1708

13

142

3,302

211

5

1708

1708

1659

%

23

40

20

12

5

0

0.36

3.63

89.81

6.090.12

87.47

4.57

7.85

0.12

48.83

51.05

0.12

n

456

734

462

210

103

0

23

7

80

1769

107

3

1712

62

188

3

886

1056

3

N

842

1414

816

407

194

1965

13

142

3,302

211

5

1965

1965

1902

%

23

37

24

11

5

0

0.35

4.07

89.97 5.45 0.15

87.12

3.16

9.57

0.15

46.11

53.74

0.15

0.239

0.877

0.047

0.250

*Demographic data from 12 week questionnaire, neonatal/infant death questionnaire, and miscarriage/stillbirth questionnaire

EXPOSURE TO INTERVENTION

Table 3.1: Exposure to intervention for healthy normal birth-weight babies

Number of women/infant dyads enrolled in the intervention arm that received home visits (newborn healthy and normal weight)

(N= women/infant dyads in intervention arm infant was alive at 12 week interview date1)

All visits

Visit within scheduled window (PN1 only)

Visits outside scheduled window (PN1 only)

AN1 (2nd trimester of pregnancy)

AN2 (3rd trimester of pregnancy)

PN1 Visit (24-48 hours after discharge)

PN2 Visit (3-4 days after discharge)

PN3 Visit (10-14 days after discharge)

PN4 Visit (3-4 weeks after discharge)


Woman/infants receiving all 7 visits

LBW babies that received: (N=LBW)

1st extra visit

2nd extra visit

n

1204

1099

957

1021

1145

1198

1274

634

49

50

N

1629

88

%

73.91

67.46

58.74

62.68

70.29

73.54

78.21

18.85

55.65

56.82

n

-

-

-

-

-

-

-

-

N

-

-

-

-

-

-

-

-

%

-

-

-

-

-

-

-

-

n

-

-

-

-

-

-

-

-

N

-

-

-

-

-

-

-

-

%

-

-

-

-

-

-

-

-

*Data from CHW reporting.

1Miscarriages, false pregnancies, stillbirths, and infant deaths were not included in the denominator after the date of miscarriage/death or false pregnancy discovered. Lost to follow up not included – only included those living who had an interview at 12 weeks.

Table 3.2: Exposure in control areas

Number women/infant dyads enrolled in the control arm that received home visits

(N= women/infant dyads in control arm where infant was alive at 12 week assessment)

All visits

AN1 (2nd trimester of pregnancy)



Woman/infants receiving all 3 visits

n

1838

1685

1674

1576

N

1865

%

98.55

90.35

89.76

45.11

*Data from CHW reporting.

1Miscarriages, false pregnancies, stillbirths, and infant deaths were not included in the denominator after the date of miscarriage/death or false pregnancy discovered. Women/infants lost to follow up are included in the denominator at all visits.

GOAL: Reduce neonatal mortality

Table 4.1: Neonatal mortality/Stillbirth

Number of neonatal deaths/stillbirths among women/infant dyads enrolled in study

(N= all women enrolled) 3957

Intervention area

(N=1821 )

Control area

(N=2136 )

p-value

Neonatal death (0-28 days after birth)

Early neonatal death (<=7 days)

Stillbirth/Miscarriage

Loss to follow-up before 28 days (after live birth)

n

21

16

50

88

%

1.15

0.80

2.75

4.83

n

21

17

68

132

%

0.98

0.88

3.18

6.18

0.603

0.948

0.420

*Data from clinic records.

Table 4.2: HIV free survival

Number of enrolled infants born to HIV-positive women by infant’s HIV status 12 weeks after birth

(N = all HIV-positive women interviewed at 12 weeks) TOTAL N= 1292

Intervention area

(N=579)

Control area

(N=713 )

p-value

Mother HIV positive and:

HIV infected at 12 weeks

Deaths before 12 weeks

HIV-free survival at 12 weeks

n

1

3

37

%

0.17

0.52

6.38

n

5

6

47

%

0.68

0.84

6.58

0.166

0.617

?

*Data from lab results database.

SO:Use of newborn healthcare services & adoption of healthy Practices/behaviors at scale

Antenatal care

Table 5.1: Antenatal care (ANC)

Percent of interviewed women that received at least 1 antenatal care (ANC) visit from a medically trained provider at a health facility (N= all women interviewed in the study- including maternal deaths)

Intervention area

(N=1708 )

Control area

(N=1965 )

p-value

Received at least 1 ANC visit at a facility

No ANC care during pregnancy

Don’t remember/Missing

n

1697

9

2

%

99.36

0.53

0.12

n

1946

16

3

%

99.03

0.81

0.15

0.278

0.548

0.548

*Questions 12.2 & 12.3 on 12 week questionnaire, neonatal/infant death questionnaire, and miscarriage/stillbirth questionnaire.

Table 5.2: Antenatal care (ANC)

Percent of interviewed women that received at least 4 antenatal care (ANC) visits from a medically trained provider at a clinic (N= all women interviewed in the study)Comment by Tomlinson, MR, Prof : We could not find the data for this

Intervention area

(N= )

Control area

(N= )

p-value

Received at least 4 ANC visits at a facility

Received less than 4 or no ANC


n

%

n

%

Intrapartum Care

Table 6.1: Place of deliveryd_119

Percent of live births delivered at a health facility (N= all women interviewed at 12 weeks who had a live birth) Total N= 3561

Intervention area

(N=1659 )

Control area

(N=1902 )

p-value

Delivered at facility

Delivered elsewhere


n

1624

35

0

%

97.89

2.11

n

1858

45

0

%

97.63

2.37

0.607

*Question 13.2 on 12 week questionnaire, neonatal/infant death questionnaire, caregiver questionnaire.

Table 6.2: Breathing Assistance d_124-129

Percent of interviewed women who had live births where breathing assistance was given at birth 1 (N= all living women interviewed at 12 weeks who had a live birth= 3556)

Intervention area

(N= 1,657)

Control area

(N=1,899 )

p-value

Breathing assistance action:

Rubbed/massaged

Dried

Mouth cleared

Nothing

Other

At least 1 breathing assistance action


n

116

7

73

424

865

1235

0

%

7

0.42

4.41

25.59

52.20

74.53

n

114

15

75

503

1005

1399

0

%

6

0.79

3.95

26.49

52.92

73.67

0.462

0.362

0.760

0.795

0.873

0.632

*Question 14.1 on 12 week questionnaire and neonatal/infant death questionnaire.

1Women asked about asphyxia management techniques used, whether or not newborn had trouble breathing/crying at birth.

Postnatal Care

Table 7.1: Postnatal check at clinic within 7 days after birth d_136

Percent of live births where baby had a postnatal check at a clinic within 7 days after birth (N= all women interviewed at 12 weeks who had a live birth)

Intervention area

(N= 1659)

Control area

(N= 1902)

p-value

Had postnatal check at clinic within 7 days of birth

Not applicable

n

1167

44

%

70.34

2.65

n

1207

56

%

63.46

2.94

0.000

*Need to check form

Table 7.2: Early breastfeeding 147-148

Percent of interviewed women with live births where newborn was breastfed within 1 hour after birth (N= all women interviewed at 12 weeks who had a live birth excluding maternal deaths) N=3556

Intervention area

(N= 1657)

Control area

(N= 1899)

p-value

Breastfed within 1 hour birth

Not given any breast milk after birth

n

567

388

%

34.07

23.42

n

609

509

%

32.22

26.80

0.105

*Question 15.2 on 12 week questionnaire and neonatal/infant death questionnaire.

Table 7.3: Exclusive breastfeeding

Percent of women interviewed 12 weeks after birth where newborn was exclusively breastfed in 24 hour period prior to survey (N= all women interviewed at 12 weeks with living baby, HIV-positive women, or HIV-negative women) N= 3494

Intervention area

Control area

p-value

Exclusively breastfed 24 hours before survey

Did not exclusively breastfeed 24 hrs before survey


By HIV-status of woman:

Exclusively breastfed 24 hrs before survey - HIV+

Exclusively breastfed 24 hrs before survey - HIV-

n

441

1888

0

133

308

N

1629

571

1058

%

27.07

72.93

23.29

29.11

n

260

1605

0

106

154

N

1865

699

1167

%

14.94

86.06

15.93

13.21

0.000

0.000

*Questions 17.2 and 17.3 on 12 week questionnaire.

IR 3: Demand for newborn services increased

Table 8.1: Birth plans

Percent of interviewed women that had a birth plan (N= all women interviewed at 12 weeks= 3668)

Intervention area

(N=1706)

Control area

(N=1962)

p-value

Women that had a birth plan with the following components:

Facility identified

Transport

Financial

Women with a birth plan with at least 1 component above

n

163

329

765

853

84

/

%

9.54

19.26

44.79

49.94

4.92

n

152

247

702

786

69

/

%

7.74

12.57

35.73

40.00

3.51

0.000

0.000

0.000

0.000

0.000

*Question 12.10 on 12 week questionnaire, neonatal/infant death questionnaire, and miscarriage/stillbirth questionnaire.

Table 8.2: Knowledge of danger signs for newborns

Percent of women interviewed 12 weeks after birth that could cite at least 2 danger signs for newborns (N = all women interviewed at 12 weeks- excluding maternal deaths but including all others =3668)

Intervention area

(N=1706 )

Control area

(N=1962 )

p-value

Women that cited the following newborn danger signs:

Convulsions

Fever/Hot body

Difficulty breathing

Difficulty sucking

Redness/Discharge around cord

Jaundice: Yellow palms, soles, or eyes

Baby is weak and soft

Weak abdominal cry

Other

Women that could cite at least 2 danger signs


n

18

575

736

146

24

104

277

99

1032

1197

245

%

1.06

33.70

44.72

8.56

1.41

6.10

16.24

5.80

60.49

70.81

14.36

n

20

601

827

168

18

67

319

85

1144

1442

294

%

1.02

30.63

42.15

8.66

0.92

3.41

16.26

4.33

58.31

73.88

14.98

0.834

0.130

0.274

0.837

0.324

0.001

0.836

0.111

0.386

0.026

0.709

*Question 25.2 on 12 week questionnaire.

Table 8.3: Newborn illnesses

Percent of live births where newborn experienced illness during the first month of life (N = all live births=3561)Comment by Tomlinson, MR, Prof : All women/ caregivers interviewed at 12 weeks with live birth

Intervention area

(N= 1659)

Control area

(N= 1902)

p-value

Newborns with the following reported symptoms:

Convulsions

Fever/Hot body

Difficulty breathing

Difficulty sucking

Redness/Discharge around cord

Jaundice: Yellow palms, soles, or eyes

Baby is weak and soft

Weak abdominal cry

OtherComment by Tomlinson, MR, Prof : No “other “option

Newborns with at least 1 symptom from above list


n

46

87

45

180

36

105

33

43

385

49

%

2.77

5.24

2.71

10.85

2.17

7.66

1.99

2.60

23.21

2.9

n

44

107

56

238

44

105

53

48

428

63

%

2.32

5.63

2.94

12.51

2.31

5.52

2.79

2.52

22.5

3.2

0.443

0.569

0.592

0.196

0.620

0.024

0.195

0.641

0.618

0.817

*Question 25.3 on 12 week questionnaire, 21.3 on neonatal/infant death questionnaire, and 21.3 on caregiver questionnaire.

Table 8.4: Care seeking for newborn illnesses

Percent of newborns experiencing illness in the first month of life and care sought (N = # newborns with at least 1 reported symptom – from Table 8.3) N= 814

Intervention area

(N=385 )

Control area

(N= 428)

p-value

Care sought from:

Public facility (clinic, hospital)

Private facility (clinic, hospital, pharmacy, doctor)

Elsewhere (incl traditional, relative or friend or other)

Care sought from at least 1 source

No care sought


n

236

42

29

90

69

9

%

61.30

10.91

7 .53

20.98

17.92

2.34

n

240

63

27

71

95

4

%

55.94 14.69

6.29

18.44

22.14

0.93

0.219

0.364

0.189

*Question 25.5 on 12 week questionnaire, 21.5 on infant death questionnaire, and 21.5 on caregiver questionnaire.

22

Annex 1: STUDY TIMELINE

Annex 3: paper queue

Insert

1.Bradshaw D, Chopra M, Kerber K, Lawn JE, Bamford L, Moodley J, et al. Every death counts: use of mortality audit data for decision making to save the lives of mothers, babies, and children in South Africa. Lancet. 2008; 371(9620): 1294-304.

2.Liu L, Johnson HL, Cousens S, Perin J, Scott S, Lawn JE, et al. Global, regional, and national causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. Lancet. 2012.

3.Chopra M, Daviaud E, Pattinson R, Fonn S, Lawn JE. Saving the lives of South Africa's mothers, babies, and children: can the health system deliver? Lancet. 2009; 374(9692): 835-46.

4.Kerber KJ, de Graft-Johnson JE, Bhutta ZA, Okong P, Starrs A, Lawn JE. Continuum of care for maternal, newborn, and child health: from slogan to service delivery. Lancet. 2007; 370(9595): 1358-69.

5.Kumar V, Mohanty S, Kumar A, Misra RP, Santosham M, Awasthi S, et al. Effect of community-based behaviour change management on neonatal mortality in Shivgarh, Uttar Pradesh, India: a cluster-randomised controlled trial. Lancet. 2008; 372(9644): 1151-62.

6.Baqui AH, El-Arifeen S, Darmstadt GL, Ahmed S, Williams EK, Seraji HR, et al. Effect of community-based newborn-care intervention package implemented through two service-delivery strategies in Sylhet district, Bangladesh: a cluster-randomised controlled trial. Lancet. 2008; 371(9628): 1936-44.

7.SANDOH. Re-engineering Primary Health Care in South Africa. . In: Health, editor. Pretoria: South African National Department of Health; 2010.

8.Aboud FE, Singla DR. Challenges to changing health behaviours in developing countries: A critical overview. Soc Sci Med. 2012.

9.Tylleskar T, Jackson D, Meda N, Engebretsen IM, Chopra M, Diallo AH, et al. Exclusive breastfeeding promotion by peer counsellors in sub-Saharan Africa (PROMISE-EBF): a cluster-randomised trial. Lancet. 2011; 378(9789): 420-7.

10.DOH. Demographic and Health Survey 2003. Pretoria: Department of Health 2007.

11.DOH. South Africa Demographic and Health Survey 1998. Pretoria: Department of Health; 2002.

12.Paim J, Travassos C, Almeida C, Bahia L, Macinko J. The Brazilian health system: history, advances, and challenges. Lancet. 2011; 377(9779): 1778-97.

Good Start III study timeline

2012

IMPLEMENTATI

ON

Jan-08Feb-08Mar-08Apr-08May-08Jun-08Jul-08Aug-08Sep-08Oct-08Nov-08Dec-08Jan-09Feb-09Mar-09Apr-09May-09Jun-09Jul-09Aug-09Sep-09Oct-09Nov-09Dec-09Jan-10Feb-10Mar-10Apr-10May-10Jun-10Jul-10Aug-10Sep-10Oct-10Nov-10Dec-10Jan-11Feb-11Mar-11Apr-11May-11Jun-11Jul-11Aug-11Sep-11Oct-11Nov-11Dec-11

2012

EVALUATION


Cost

2012

OUTPUTS


Publications

planned

LINK TO

NATIONAL

SCALE

2012


CONTRACTS

SNL contract and

reports

CDC contracts

Community

Impact and

process

2009

Enrollment period

FGDs

Clinic

200820092010

2008200920102011

20102011

Running costs

Home visits (2 Antenatal, 5 Postnatal)

2011

FGDs

Community

Randomization

& mapping

Quality

improvement

implementation

(PDQ)

Building

Support

Set up costs

SNL contract with MRCExtension - no cost to Dec 2012

QITs

begin

2008

2008200920102011

Data collection (Data collection point: 12 weeks postnatal)

Data cleaningAnalysis & write up

Bridging

Activities

SNL Indicator Enrollment ends 1Dec2010Analytic plan Prelim analysisSNL End Report Mar 2012CHW training2010 Annual

report due

2009 Annual

report due

2008 Annual

report due

Protocol paper submittedImpact CEApaperFinal enrollmnt data incl in Q3 reportStakeholder summitPDQ training workshopFGD AnalysisFGD AnalysisBridging workshopAnalysis workshopFinal training manuals2011 Annual report dueCDC funding stopsBaseline done Sep-Nov07

· Web viewSAVING NEWBORN LIVES. South Africa . Good Start III . End of Study Report . September....

Documents

Transcript of · Web viewSAVING NEWBORN LIVES. South Africa . Good Start III . End of Study Report . September....