THE UNITED REPUBLIC OF TANZANIA MINISTRY OF HEALTH AND SOCIAL WELFARE

THE UNITED REPUBLIC OF TANZANIA MINISTRY OF HEALTH AND SOCIAL WELFARE Baseline survey on quality of paediatric care in Tanzania

Published by the Ministry of Health and Social Welfare, Dar es Salaam, The United Republic of Tanzania. 2011 Ministry of Health and Social Welfare, The United Republic of Tanzania Any part of this document may be reproduced in any form without the prior permission of the publisher provided that this is not for profit and that due acknowledgement is given. Any reproduction for profit must be made with the prior permission of the publisher. Copies of this report may be obtained from: The Permanent Secretary Ministry of Health and Social Welfare P.O. Box 9083 Dar es Salaam Tel: 255 22 212021 Fax: 255 22 2139951

Table of contents Acknowledgments... i Abbreviations... ii Executive summary...iii Background...iii Objective...iii Methods...iii Results...iv Recommendations... v Conclusion...vi 1. Introduction... 7 2. Background... 8 2.1 Tanzaniaʹs health system... 8 2.2 Quality of care in children... 9 2.3 Objective of the assessment... 11 3. Methods... 12 3.1 Scope of the survey... 12 3.2 Performance criteria... 12 3.3 Grading performance... 12 3.4 Assessment teams... 14 3.5 Data entry and analysis... 14 3. Statistical analysis... 15 4. Results... 15 4.1 Administrative review... 1 4.1.1 Overall performance... 1 4.1.2 Hospital support... 17 4.2 Clinical assessment... 25 4.2.1 Paediatric ward... 25 4.2.2 Monitoring... 25 4.2.3 Infection prevention and control... 2 4.2.4 Care for children by qualified staff... 27 4.2.5 Case management... 28 5. Newborn care... 3 5.1 Delivery room/ward (nursery)... 37 5.2 Delivery of newborn care... 37 5.2.1 The neonatal resuscitation variables scored were:... 37 5.2.2 For promotion of early breastfeeding and bonding the variables were:... 37 5.2.3 Clean delivery and newborn care variables were:... 38 5.2.4 Prophylaxis variables were:... 38 5.3 Sick newborn/neonate care... 38. Further data analysis of the survey findings... 39 7. Discussion... 45 7.1 Conclusion... 48 8. Recommendations at regional level... 48 9. Recommendations at national level... 48 10. References... 49 11. Annexes: Tables showing the overall hospital performance by region... 50

Acknowledgments The Ministry of Health and Social Welfare (MOHSW) would like to extend its sincere appreciation to all the people who contributed to the development of this report. It is not possible to name all who have played a part in data collection and data entry; however the Ministry would like to mention a few whose particular dedication to this process made this report possible. Acknowledgements go first to all those individuals who participated in the development of the data collection tool and the check list, as well as those who assisted with the data collection. In the Reproductive and Child Health Section, MOHSW, we thank Drs Azayo Mary, Bundala Felix, Mr Meena John, Drs Msemo Georgina and Rusibamayila Neema. The MOHSW is especially grateful to the WHO Country Office as well as WHO HQ for the financial and technical support provided, especially to Drs Hill Sue, Iriya Neemes and Shija Rose. Finally, the MOHSW would like to thank the Principal Investigator and author of this report, Dr Kitundu Jesse from Muhimbili National Hospital, Department of Paediatrics, for leading this important work to its completion. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page i

Abbreviations APC ARI AS BF CAS CPR CWS DPT3 EmOC FBO IMCI IMR IPD IPTP HSR MDG MNCH MOHSW NMR OPD ORS ORT PAT PHS PMTCT PNC ResoMal RCHS RCM SAM SOP SM SPM SNM TB TS TDHS U5MR WHO Acute patient care Acute respiratory infection Administrative score Breastfeeding Clinical assessment score Contraceptive prevalence rate Childrenʹs ward score Three doses of diphtheria, pertussis and tetanus vaccine Emergency obstetric care Faith based organization Integrated management of childhood illness Infant mortality rate Inpatient Department Intermittent preventive treatment for malaria during pregnancy Health System Reforms Millennium Development Goal Maternal, newborn and child health Ministry of Health and Social Welfare Neonatal mortality rate Outpatient Department Oral rehydration solution Oral rehydration therapy Paediatrics Association Tanzania Public health score Prevention of mother to child transmission of HIV Postnatal care Rehydration solution for severe malnutrition Reproductive and Child Health Services Referral care manual Severe acute malnutrition Standard operating procedure Standard met Standard partially met Standard not met Tuberculosis Treatment score Tanzania Demographic Health Survey Under five mortality rate World Health Organization Baseline Survey on Quality of Paediatric Care in Tanzania -- Page ii

Executive summary Background This report outlines the preliminary results of an ongoing assessment survey on the quality of paediatric care conducted by the Ministry of Health and Social Welfare in the United Republic of Tanzania. Mainland Tanzania has 21 administrative regions and 113 districts. The country has a pyramidal referral structure of health care with public and private dispensaries, health centres and district, regional and national hospitals managed by the Government and by non governmental and faith based organizations. The survey reported here covers 9 hospitals managed by the Government and faith based organizations across regions of Mainland Tanzania. Millennium Development Goal 4 aims for a reduction in child mortality. Quality of care is an important factor in reaching this goal. While under five mortality has decreased in Tanzania, the rate in hospitals remains unacceptably high with 75% of these deaths occurring in the first 24 to 48 hours after admission. Most of these deaths are preventable and the application of appropriate measures, such as proper assessment, treatment and care, could reduce the number of deaths significantly. Objective The goal of the survey is to establish baseline data and identify gaps to be addressed in order to improve the quality of paediatric care in Tanzanian hospitals. The specific objectives of the survey are to: Methods assess the administrative and logistics support in the provision of care for sick children; assess the knowledge and skills of health workers in the management of care for common childhood illnesses; assess the availability of essential medicines, supplies and equipment at health facilities necessary for the provision of quality care; support the establishment of a system for improving the care of children in the respective facilities. The World Health Organization s generic tool for assessing the quality of hospital care for children was adapted to collect data for this survey. The WHO Pocket Book of Hospital Care for Children and an adapted referral care manual were used as standards for assessing case management. Data were collected in hospitals and health centres through observation and interviews by teams comprising three trained assessors from other facilities. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page iii

The five areas assessed were administration, paediatric wards, clinical assessment, public health and treatment. Each area had key indicators to be measured, including the availability of essential medicines, adequate qualified staff, a separate paediatric ward, accurate clinical assessment and treatment of pneumonia, diarrhoea and malnutrition, reassessment after admission, and promotion of early breastfeeding. The indicators used for scoring were divided into two categories. First, clinical tasks and standards with a strong bearing on the care of the child, the absence of which could be life threatening. They include the availability of emergency medicines and accurate clinical assessment and treatment for dehydration. The second category, essential tasks and standards, covers the availability of laboratory facilities for culture testing and failure to recognize a skin infection. Scoring was categorized as: standard met, standard partially met and standard not met. This method of scoring allows comparison between Government and non Government hospitals, between hospitals within the region and also between regions. Results The results varied greatly across the measured variables and from region to region. Overall the assessment results are poor, particularly in clinical assessment. They show that of the 82 variables measured none of the hospitals scored more than 75%. Of the 9 hospitals assessed, 42 0.9% had total scores of less than 50%. The highest scoring regions were Mwanza and Mbeya, while the lowest scoring regions were Lindi and Mtwara. Government hospitals scored lower than faith based hospitals in most of the areas assessed. Emergency care, diarrhoea assessment, management of severe malnutrition and newborn care were among the worst scoring variables. Compared with the assessment of clinical conditions, HIV/AIDS testing, counselling and treatment performance were high scoring. The presence and availability of appropriate and adequate human resources scored poorly. Less than 25% of the facilities fully met the standards for qualified staff providing care to children, with over 50% of facilities partially meeting the standards. The availability of standard treatment guidelines, essential medicines and equipment were among items assessed as part of the administration and paediatric ward assessments. Findings show a lack of adequate and updated treatment guidelines in all hospitals. The availability of essential medicines as per the Essential Medicines List was poor; with just over one third of hospitals having the medicines on the list. Less than 25% of the hospitals included in the survey had essential equipment and supplies. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page iv

Recommendations A number of recommendations were made: Revise and focus the National Reproductive and Child Health Strategic Plan with a priority on addressing areas that scored poorly in the survey. For example, assessment and records in emergency and paediatric wards are essential to initiate changes in paediatric quality of care. Ensure all health facilities have updated clinical standard treatment guidelines for children and essential medicines lists, as they will strengthen the knowledge and clinical skills and hence the quality of care. Improve the availability of essential medicines, especially emergency medicines, supplies and hospital and laboratory equipment. Allocate trained staff in paediatrics to improve clinical assessment and diagnosis. Implement a continuous education programme for medical staff. Training should be decentralized to the regional level to allow programmes to address the differing needs from region to region. The Paediatric Association of Tanzania should take the lead in coordinating continuous education programmes, with referral hospitals and medical teaching institutions responsible for conducting the training. Introduce sharing learning sessions at the regional level for facilities to learn from each other. Take into consideration the inadequate number of paediatricians and neonatologists. Clinical officers and midwife nurses should be allowed to train in general paediatrics and neonatology combined. This will reduce the current shortage of health workers with the necessary knowledge and skills for providing quality paediatric and newborn care at district and regional levels. Improve the infrastructure for paediatric and neonatal wards in all hospitals by having separate outpatient and inpatient areas for children and newborns. Strengthen monitoring, support and mentorship systems in order to achieve good quality of care. As good quality care is costly, more Government funds should be allocated for improving the quality of care. Increase advocacy for quality of care at all levels: policy makers, administrators, health workers, those in pre and post medical training institutions and non health workers. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page v

Conclusion The overall performance of hospitals was poor in almost all areas. There were variations from region to region and between Government and non Government hospitals. Clinical assessment of children admitted to paediatric wards is very poor and is associated with misdiagnosis and inappropriate treatment. While the child mortality rate has gone down, more effort is needed to reach Millennium Development Goal 4. Steps need to be taken nationwide to improve the situation. These measures should be based on a revived and more focused National Reproductive and Child Health Strategic Plan with priority given to those areas that scored poorly in this survey. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page vi

1. Introduction The United Republic of Tanzania is striving to improve the quality of its paediatric care with the aim of substantially reducing child mortality. Efforts to improve maternal, newborn and child heath service delivery have already resulted in considerable gains. Under five child mortality declined from 137 deaths per 1000 live births in 199 to 81 deaths per 1000 live births in 2010. 1 Tanzaniaʹs main goal is to improve the quality of paediatric care at all health facilities in an effort to reach Millennium Development Goal (MDG) 5, which calls for a reduction in the mortality of children under age 5 by 2015. In 200, Tanzania drew up a 10 year plan to establish a framework for the integration of maternal, neonatal and paediatric care. In 2008 it became known as the National Reproductive and Child Health Strategic Plan ( One Plan ). In accordance with its strategy, the Ministry of Health and Social Welfare (MOHSW) is seeking to establish baseline data on quality of paediatric care that will allow better planning and more targeted interventions. The Ministry conducted an assessment of the quality of paediatric care in public and private hospitals and health centres throughout Mainland Tanzania. Sustained availability and access to medicines suitable for children is an essential component of any strategy to improve paediatric care. In May 2007, the World Health Organization (WHO) World Health Assembly passed a resolution to strengthen and support activities to make medicines more readily available to children and to promote the development of evidence based treatment guidelines to ensure that drugs are used appropriately. Following this, WHO initiated the Better Medicines for Children (BMC) project in 2009, with funding from the Bill and Melinda Gates Foundation. The goal of the project is to improve access to essential medicines for children by addressing issues of availability, safety, efficacy and price. One of the four objectives of the BMC project is to improve access to essential medicines for children in priority countries by promoting their inclusion in national essential medicines lists (EMLs), treatment guidelines and procurement schemes; working with drug regulatory authorities to expedite regulatory assessment of essential medicines for children; and 1 Tanzania Demographic and Health Survey 2009 2010. United Republic of Tanzania, 2010. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 7

developing measures to monitor and manage the prices of these medicines. Tanzania is one of the countries participating in the project. This assessment will assist the projectʹs aims by indicating which hospitals and regions lack access to essential medicines for children and also lack guidelines for their appropriate use, which are important components for ensuring quality of care. Results presented in this report cover 9 hospitals from 12 regions of Mainland Tanzania (Coast, Dodoma, Iringa, Kagera, Kigoma, Lindi, Mbeya, Morogoro, Mtwara, Mwanza, Shinyanga and Tabora). The hospitals were assessed between July 2009 and February 2010. The survey is ongoing in the remaining nine regions (Arusha, Dar es Salaam, Kilimanjaro, Manyara, Mara, Rukwa, Singida, Ruvuma and Tanga). 2. Background The United Republic of Tanzania is a union between Tanganyika (Mainland Tanzania) and Zanzibar. Mainland Tanzania is divided into 21 administrative regions comprised of 113 districts with around 10 342 villages and 133 councils. Primary health care forms the basis of health care services, which have a pyramid structure. Both public and private providers work in dispensaries, health centres and at least one hospital at the district level. There are 479 dispensaries and 481 health centres throughout the country. Notably, about 90% of the population lives within 5 kilometres of a primary health facility. The Government owns 55 district hospitals, while 13 district hospitals are owned by faithbased organizations. Mainland Tanzania has 8 hospitals at first referral level (owned by the Government, parastatals1 and the private sector), 18 regional hospitals that function as referral hospitals for district hospitals, and eight consultancy and specialized hospitals. Government staffing norms for health facilities exist, but are not fully met. When comparing these norms to staffing levels in all health facilities, only 35% of positions are filled by qualified health workers. This constitutes a severe human resource crisis. 2.1 Tanzania's health system Referral pathway Tanzania has categorized health facilities into three main groups: dispensaries health centres hospitals Within the Government system, hospitals are further categorized as: 1 An organization or industry having some political authority and serving the state indirectly. (Oxford English Dictionary, 11 th edition, 2009.) Also defined as: A government owned corporation, state owned company, state owned entity, publicly owned corporation, government business enterprise, or parastatal is a legal entity created by a government to undertake commercial activities on behalf of an owner government. (Wikipedia, http://en.wikipedia.org/wiki/government owned_corporation.) Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 8

district hospital regional hospital national hospital consultant or tertiary hospital special hospital (e.g. Mirembe, Ocean Road Cancer Institute). The referral pathway begins at the dispensary, and then leads to a health centre. Patients are referred to hospital by the health centre. In some cases, depending on the diagnosis or availability of beds a child may be referred directly to a district hospital. A child may also be referred from a district hospital directly to a consultant hospital without passing through a regional hospital. However, a dispensary must not refer beyond the district/regional hospital. As referral care must involve follow up, feedback is sent to the referring facility with clear instructions on management and the date of any follow up visits needed. 2.2 Quality of care in children Generally, the quality of paediatric care has improved over the past 10 15 years. For example, national Integrated Management of Childhood Illness (IMCI) coverage is 93.8%; measles immunization is 98.0%; and national oral rehydration treatment (ORT) coverage, 70.0%. Because of these gains, significant progress has been made in reducing child mortality, although neonatal mortality still remains high, at an average of 32 deaths per 1000 live births. This accounts for 47% of infant mortality, at a rate estimated at 8 deaths per 1000 live births. The figure below shows trends in childhood mortality over the past 10 15 years. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 9

Figure 1: Trends in early childhood mortality rates 1 Red = Millennium Development Goals (MDGs) Blue = National mortality rate (NMR) Orange = Infant mortality rate (IMR) Purple = Under 5 mortality rate (U5MR) While overall child mortality has decreased, in hospital settings deaths among children remain unacceptably high. Seventy five per cent of these deaths occur within the first 48 hours of admission. Most of these deaths are preventable and could be reduced significantly if appropriate measures were taken including prompt referrals, appropriate assessment, treatment and care when the patient arrives at the facility. Studies 2,3,4 have shown that about 10% of children seen at a first level health facility require a referral. Unfortunately, only 30% of these children are given a referral. Of those who do receive a referral, less than half are able to attend the referral appointment due to difficulty with transport, high cost, or perceptions of poor attitudes by health workers and poor quality of care. 1 2 3 4 Trends in early childhood mortality rates. Tanzania Demographic and Health Survey (TDHS), 2010. Masanja H, de Savigny D, Smithson P, et al. Child survival gains in Tanzania: analysis of data from demographic and health surveys. The Lancet, Vol. 371, Issue 920, 127 1283, 12 April 2008. Tanzania Demographic and Health Survey 2010. United Republic of Tanzania. Acute Paediatric Care Manual 2005. Muhimbili National Hospital, Department of Paediatrics and Child Health. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 10

Quality of care at the referral facility remains insufficient, with a significant number of deaths (50%) occurring within the first 24 hours of admission. Similarly, inpatient monitoring remains poor. 1 In 2007, an acute patient care (APC) concept was established at Muhimbili National Hospital (MNH) that improved paediatric quality of care. As a result, child mortality rates in general paediatric wards at MNH fell from 17% to below 10% in 2008 2009. In the first half of 2009, the APC concept was adapted and incorporated into the National Reproductive and Child Health Strategic Plan for implementation and supervision. This plan was followed by training seminars for all staff doctors and nurses working in paediatric wards in all regional hospitals and referral hospitals (except Kilimanjaro Christian Medical Centre, but including Mnazi Mmoja Hospital in Zanzibar). After completion of the training, both in Mainland Tanzania and Zanzibar, supportive supervision and mentorship was conducted in all 2 regional hospitals, including Mnazi Mmoja, to assess the establishment and functionality of acute paediatric units. After national implementation, it was found that only 25% of hospitals in the area being assessed had established APC units and less than 10% were fully operational. The main reason given for poor implementation was lack of adequate staff at all facilities. 2.3 Objective of the assessment The objective was to establish baseline data and identify strengths and weaknesses that may be used to plan interventions for improving paediatric quality of care in Tanzania. The survey is designed to assess the: administrative and logistic support in the provision of care for sick children; knowledge and skills of health workers in the management of care for common childhood illnesses; availability of essential medicines and supplies and equipment at health facilities necessary for provision of quality paediatric care. The assessment supports the establishment of a system for improving care of children in health facilities. 1 Seven countries assessment report, IMCI baseline survey, JMP Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 11

3. Methods 3.1 Scope of the survey This project required continuous assessments of health facilities using adapted national assessment tools, followed by an improvement plan and evaluation. It also included supportive supervision and on site capacity building. The assessment, using an adapted WHO assessment tool Assessment of the Quality of Care for Children in Hospitals involved all systems within health facilities in all regions of Mainland Tanzania. Areas that were assessed on the care of children under five years included: organization of health facilities; outpatient care; emergency care; emergency paediatric wards; diagnostic areas; dispensing pharmacies; kitchens; laundry facilities; and other relevant departments. Emphasis was placed on the assessment of the skills and knowledge of health workers in the management of emergencies, common clinical conditions and inpatient monitoring of seriously ill children. Newborn care was also assessed at all health facilities. The assessment covered provision of services during and after working hours, during the night, weekends and holidays. At the end of the survey, health facility management and staff were debriefed by assessment teams. 3.2 Performance criteria The performance of health facilities and individual health workers was measured against agreed standards. The standards have been adapted from WHO standards of care and also took into consideration requirements and recommendations by different disease programmes. The WHO Pocket Book of Hospital Care for Children 2005 and adapted Referral Care manuals (RCMs), and the Integrated Management of Childhood Illness for High HIV Settings 200 were used as standards for assessing case management. 3.3 Grading performance The criteria used to grade performance were defined as the baseline for measuring the subsequent improvement of care. In addition, the performance of individual tasks was considered for provision of feedback and planning. The tasks used as the basis for grading performance were divided into two categories. A: Critical tasks or standards that have a strong bearing on the outcome of care of children, such as the availability of emergency drugs, failure to recognize dehydration, etc. These are identified with an asterisk in the assessment forms. B: Essential tasks or standards. Failure to fully meet these standards or complete these tasks does not significantly impair care and is not life threatening, e.g. availability of culture facilities, failure to recognize skin infection, etc. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 12

The scoring was categorized as follows: STANDARD MET: All critical tasks were completed to standard. STANDARD PARTIALLY MET: All critical tasks were completed to standard, but all or some of the essential tasks were missed. STANDARD NOT MET: All critical and all essential tasks were not met. The tool used during the assessment of tasks is summarized below. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 13

Table 1: Summary of the health facility assessment tool Assessment area General information to be collected Specific individual service Delivery areas Reception Emergency management Consultation room Drug supplies and equipment Assess laboratory service provision Paediatric ward Newborn care Information needed Types of service delivery areas Adequacy of facility and staffing Facility utilization and bed occupancy rates Morbidity and mortality Observe/assess activities carried out in these areas: Distance between services Availability of essential equipment and drugs Round the clock staff coverage Quality of medical consultation and treatment Adequacy of: Spectrum of tests performed and their quality Equipment and reagents Staff coverage Staff coverage of wards Staff knowledge about appropriate care of sick children Staff attitude Emergency treatment and procedures Presence of patient isolation areas Availability of necessary equipment and drugs Patient monitoring Feeding of severely malnourished children Role of mothers or care givers in child care Counselling mothers or care givers at time of discharge Patient follow up Communication with other facilities Children s play and other stimulating facilities Adequacy of discharge procedures Warmth and temperature control Infection prevention Nourishment and breastfeeding Resuscitation Respiratory support 3.4 Assessment teams Eighteen people were recruited at the regional level for the assessment teams, drawn from district and regional level health facilities. Orientation of the teams was carried out by senior technical staff from a national quality improvement team. The teams, which each consisted of three people, conducted assessments in facilities other than their own. At the end of the assessment, feedback was given to health facility management and staff. Following the debriefing, the teams reconvened at the regional level to share their findings and agree on areas for improvement. 3.5 Data entry and analysis The information obtained from the regions was sent to the Reproductive and Child Health Service (RCHS) Unit at MOHSW for data entry and analysis. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 14

3. Statistical analysis A standard scoring system was used to assess the quality of care in different hospitals. If the facility was seen to have met the standard, it scored 3; if it partially met the standard it scored 2 and if the standard was not met, it scored 1. For statistical analysis, these scores were converted to: 0 instead of 1, 0.5 for 2 and 1 for 3 to allow the use of the statistical software SPSS version 15. The scoring was done such that, if during the assessment the facility was seen to have met the standard it scored one (1). If the standard was partially met, it scored 0.5 while if the standard was not met it scored 0. Scores were summed across the specific process, e.g. (hospital support) and also across all processes, and the summed total scores were categorized into 3 groups i.e. low (total score < 2), moderate (total score above 2 and below 4) and high quality above 4) as presented in some of the tables on performance. Cross tabulations were done by region to show any differences between hospitals in the same region and also between regions. A further analysis of comparative performance of hospitals and regions was carried out using the 25 percentile as a cut off point for performance across all 9 hospitals in 12 regions. A score of 1 was given to the hospitals that performed 25 percentile. Five domains were constructed from the individual assessment items, using five key indicators that were scored and totalled. These were: Public health score = 7 Administration score = 11 Paediatric (children s) ward score = 5 Clinical assessment score = 11 Treatment score = 10 Cross tabulations were done by domain to reveal the differences in performance between hospitals within the same region or in other regions and Spearman rho correlation coefficients were calculated using Stats Direct (version 2.7). 4. Results The results present the overall performance in percentages in terms of administration, clinical assessment, paediatric (childrenʹs) ward, treatment and public health key indicators at national and regional levels. Bar chart results are presented in three colours: GREEN, meaning the standard was met, (critical and essential tasks were performed); YELLOW, the standard was partially met (all critical and some essential tasks were performed); and RED, the standard was not met (both critical and essential tasks were not performed). Detailed results for each chart are in the tables in the annexes, showing overall hospital performance by region. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 15

4.1 Administrative review 4.1.1 Overall performance Sixty nine hospitals in 12 regions of Mainland Tanzania were assessed. There were two referral hospitals (Bugando and Mbeya), 14 private (non governmental), 12 regional and 41 public hospitals. Geographically, these hospitals were located all over Mainland Tanzania. Hospitals in the North Eastern Zone, however, were not assessed during the survey period (see the map above). Findings: Scores of the variables for individual sections were added up to establish overall scores. The total score based on 82 variables from each facility assessed, as reflected in the tables, ranged from a minimum of 14.5 (17.%) to a maximum of 1.0 (74.3%) with one hospital scoring each extreme. None of the hospitals scored more than 1.5 (75%.) Only six hospitals scored 41 or more ( 50%) and 42 (0.9%) hospitals scored less than 41 (50%) of the total variable scores. Twenty one (30.4%) hospitals had missing variable values in two or more sections and they were not included in the final analysis. Table 2: Summaries of the overall performance of hospitals, by region REGION TOTAL SCORES (82) 100% <50% 50%+ Morogoro 5 (Berege, Kilosa, Mahenge, 1 (Morogoro) St Francis, Turiani) Iringa 3 (Iringa, Makete, Njombe) 0 Mbeya 3 (Chunya, Ileje, Kyela) 2 (Mbeya Rufaa, Mbozi) Coast 5 (Kisarawe, Mafia, 1 (Bagamoyo) Mukuranga, Tumbi, Utete) Lindi (Kinyonga, Liwale, 0 Nachingwea, Nyagao, Ruangwa, Sokoine) Mtwara 4 (Ndanda, Ligula, Newala, 0 Tandahimba) Kigoma 5 (Heri Mission, Kabanga, 0 Kasulu, Kibondo, Kigoma Regional) Tabora (Igunga, Kitete, Ndala, 0 Nzega, Sikonge, Urambo) Kagera 1 (Nyakahanga) 0 Shinyanga 2 (Kahama, Shinyanga) 0 Dodoma 5 (Kongwa, Kondoa, 0 Mpwapwa, Mkoani, Mvumi) Mwanza 2 (Geita, Sekotoure) 2 (Bugando, Sengerema) Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 1

Figure 2: Hospital performance, by region Key: Green bars show the overall number of hospitals with satisfactory performance >50%. 4.1.2 Hospital support The assessment of hospital support involved six administrative items: availability of adequate and updated treatment guidelines; performance reviews; transport to referral; availability of essential medicines (in the paediatric ward and the emergency area); availability of essential laboratory tests; availability of essential equipment and supplies. 4.1.2.1 Availability of adequate and updated treatment guidelines Guidelines required were: access to a Referral Care Manual (RCM); IMCI chart booklets and hospital pocket book; Job Aids for children different charts, e.g. anthropometric measurement charts for interpretation, dehydration assessment charts, an algorithm for the ABC concept, a motherʹs card for feeding a child up to 5 years of age), and the WHO Pocket Book of Hospital Care for Children guideline; availability of standard operating procedure (SOPs) for neonates in the resuscitation area. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 17

These guidelines specifically recommended medicines for common illnesses including essential medicines. Findings: There was an alarming lack of adequate and updated treatment guidelines in all regions. Only one (1.5%) hospital Morogoro Regional Hospital met the standard and 25 (38%) hospitals partially met the standard. The majority of hospitals, 40 (0%) did not meet the standard. In Mtwara and Dodoma regions, none of the hospitals met the standard (score of zero) for guidelines. Among Government, non government and private hospitals, 0% did not have access at all to the required guidelines. Of the Government hospitals, 40% partially met the standard, as compared to non governmental and private hospitals where only 33% partially met the standard. Figure 3: Summary of the availability of standard treatment guidelines in facilities, by region Key: Green standard met; Yellow standard partially met; Red standard not met. GOVRT = Government, NOGRT =Non government 4.1.2.2 Performance reviews Performance review indicators covered: if a database of patients existed in the hospital, including records of all paediatric deaths; if regular mortality meetings were conducted to review paediatric deaths and if other staff meetings were held; quality of care by reviewing staff knowledge and skills with regard to patient care (i.e. assessment, treatment, monitoring and patient flow) with relevant staff involved in the reviews. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 18

Findings: Most of the hospitals did not satisfy the standards required for performance review. A total of hospitals were analyzed instead of 9 as 3 hospitalsʹ data were missing from the data sheet, with no explanation given. Only six hospitals (9%) met standards for monitoring, quality of care and staff knowledge; other tasks were performed well. Of these six hospitals, two were in Mbeya region (Mbozi and Tukuyu), two in the Coast region (Bagamoyo and Utete), one in Iringa (Njombe) and one in Morogoro (Kilosa); 22 (33%) of hospitals partially met and 38 (58%) did not meet the standard. Out of 18 NGO/private hospitals that were assessed in this category, only nine of these, 13.% of the total number of hospitals, partially met the standard. For Government facilities, 13 (19.%) also only partially met the standard. The worst performing region was Shinyanga where none of the hospitals met any of the standards, whereas in Morogoro region this was the case with only one hospital. Figure 4: Distribution of all hospitals with regard to findings on performance reviews, by region 4.1.2.3.1 Availability of medicines on the Essential Medicines List A number of issues were assessed: whether the Essential Medicines List was available at health facilities; whether essential medicines were available in the paediatric ward and emergency area and immediately accessible (a Category A critical requirement); conditions of storage such that old stock is used first before its expiry date; stock taking and supply systems were in place; proper handling of medicines by nurses. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 19

Findings: Thirty six percent of hospitals had medicines that were required on the Essential Medicines List. In (29) 42% of hospitals standards were not met. Morogoro, Lindi, Mtwara and Mwanza regions were a little better compared to other regions. Half of the hospitals in these regions met the standards. The worst performing regions were Iringa, Coast, Tabora and Shinyanga, which did not meet the Category A requirement. In regard to the availability of essential medicines, NGO/private hospitals were better off (>50%) as these hospitals had the required essential medicines; however, only 15 out of 51 (29%) of Government hospitals were satisfactory in this respect. Figure 5: Distribution of availability of essential medicines, by region 4.1.2.3.2 Availability of emergency essential medicines and supplies This item provided for the assessment of: Findings: the availability of essential medicines for emergency care (whether they were readily available). This was a Category A (critical) requirement. Only 19 (29%) hospitals met the standards, 33 (48%) of the hospitals failed to meet them. The worst performers were in the Iringa and Coast regions where Category A medicines and supplies were not available. Morogoro, Mbeya, Kagera and Mwanza performed much better than other regions. NGO hospitals performed a little better than public hospitals: 12 (%) partially met the standards compared to only 22 (47%) of public hospitals. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 20

Figure : Number of hospitals with availability of essential medicines for emergency care, by region 4.1.2.4 Essential equipment and supplies The assessment mainly focused on the availability of equipment and supplies on the essential equipment and supplies list at health facilities. The Category A (critical) requirement was for essential equipment to be safe and in working order. Findings: A total of 7 hospitals from 12 regions were assessed; two were dropped due to a lack of data. Fifteeen (22%) of the hospitals (public/ngo/private) had essential equipment and supplies available as per the standard. Only one of 1 (.25%) NGO/private hospitals met the required standard for availability of essential equipment and supplies; whereas 14 out of 51 (27.5%) Government hospitals met the required standard. Half of the hospitals satisfactorily met the standard in the Mwanza and Lindi regions. However, 1 (24%) did not satisfy the standard, meaning that equipment and supplies were not adequate for children in Morogoro, Kigoma, Kagera and Dodoma. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 21

Figure 7: Number of hospitals with availability of essential equipment and supplies, by region 4.1.2.5 Availability of essential laboratory support for emergency care Variables for this item were: Findings: the availability of emergency investigations; timely return of results for diagnosis and treatment; whether the cost of laboratory investigation was an impediment to the required management. Thirty three (49%) out of all the hospitals met the standards; the Mwanza and Mbeya regions performed the best. However, 20 (30%) of all hospitals did not meet the standards; in Mtwara, Tabora and Kagera regions, half of the hospitals failed to do so. NGO hospitals performed better in this area with 10 (5%) hospitals meeting standards as compared to only 23 (47%) of public hospitals. Five (28%) of the NGO and 10 (31%) of public hospitals did not meet the standards. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 22

Figure 8: Availability of essential laboratory support for emergency care, by region 4.1.2. Availability of essential equipment for emergency care The survey assessed: Findings: the availability, safety and working order of essential equipment, including whether it was on the recommended essential equipment list for emergency care. The majority of hospitals (54%) did not meet the standards. Only 12 (18%) hospitals satisfied requirements: two were in Morogoro, two in Mbeya, one in the Coast region, four in Mtwara, and three in Mwanza. The Mtwara region performed well with only one hospital not meeting the standards. However, in the Iringa, Lindi, Tabora, and Dodoma regions, the majority of hospitals fell short of the standards. For this parameter, NGO hospitals performed better than public hospitals: 22% versus 1%. Twenty eight (57%) public hospitals did not meet the standards, compared to only eight (44%) NGO hospitals. The same assessment tool that was used for public hospitals and for NGO hospitals. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 23

Figure 9: Distribution of hospitals with regard to availability of essential equipment, by region 4.1.2.7 Documentation of patient records is in place The assessment on hospital documentation examined the following: Findings: the inpatient register; patient files; ward round books; availability of monthly disease summary reports; medical record keeping in general; paediatric care (supportive care charts e.g. feeeding, medication, vital signs are recorded correctly and available); whether information could be retrieved when needed. Only 25 (3%) of hospitals met required standards while nine (13%) fell below. In Tabora and Kagera regions 33.3% of the hospitals met the standard while in Lindi region none of the hospitals performed satisfactorily. With regards to the whole system for documentation of patients records, NGO hospitals performed better than public hospitals with 10 (5%) out of 18 meeting the standard, compared to only 15 (29%) out of 51 public hospitals. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 24

Figure 10: System for documentation of patients' records, by region 4.2 Clinical assessment The Referral Care Manual (RCM), which is based on the Integrated Management of Childhood Illness (IMCI), was adopted by the Government in 2005. All the clinical variables that were assessed were taken from this manual, which is believed to be widely available, and implemented in all hospitals. RCM defines a framework for evaluating and improving standards of care. 4.2.1 Paediatric ward Several important variables for quality of care were surveyed but have not been reflected in the data collected. Missing variables were: age groups admitted in paediatrics and number of staff allocated (doctors, nurses and medical attendants). In terms of patient care, these are variables that play a vital role in outcomes. Only six variables were included in the analysis: location of the ward; if site easily accessible; allocation of seriously ill children within the ward; infection prevention and control; accommodation for the mother/care giver; documentation of patients; number of trained staff. 4.2.2 Monitoring This section focused on seriously ill children and covered four areas: receiving close attention, reassessment of admitted children, nursing care and a paediatric audit system in place. Close proximity of nursing staff to seriously ill patients may ensure that these children Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 25

are cared for in a section where they receive close attention, and care is a Category A (critical) requirement. From the four areas of the assessment the following variables were include in the analysis: Findings: proximity of seriously ill child to nursing staff; presence of monitoring charts containing all patients details; daily re assessment of newly admitted and seriously ill patients by both doctors and nurses; availability of qualified nurse 24 hours a day and medicines provided according to plan; availability of hospital paediatric care audit; availability of all children s files for audit and whether the information sought was used for improving quality of care. Almost half (49%) of the hospitals performed unsatisfactorily (standard not met) in patient monitoring. The Coast and Mtwara regions had the worst performance as no hospital met the standard. Only 22 (32%) hospitals satisfied this Category A requirement. Nearly half of these hospitals were in Morogoro, Mbeya, Dodoma and Mwanza; 11 (21%) met the standard partially and 25 (49%) did not meet the standard. Out of 9 hospitals assessed, 18 (2%) were NGO hospitals. Seven (10.1%) of NGO hospitals met the standard. Two (2.9%) partially met the standard. 4.2.3 Infection prevention and control Infection prevention and control are important for preventing cross infection during hospitalization. Nosocomial acquired infections are very difficult to manage due to high antibiotic resistance. The assessment focused on staff behaviour, namely: Findings: hand washing before and after procedures and examinations; ward cleanliness; disposal of sharps in a special safety box. Performance was generally poor as only 30 (43%) of hospitals met standards and 4% partially met the standards. In the Lindi, Kigoma and Mwanza regions, half of the hospitals met the benchmark, whereas in Morogoro, Iringa, Coast and Shinyanga less than one third of the hospitals did so. There were poor standards of hygiene in general. NGO/private hospitals performed better than public hospitals in infection prevention and control. While 4 (7%) of the hospitals met the standards, only 24 (35%) of the public hospitals did so. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 2

Figure 11: Infection control performance, by region 4.2.4 Care for children by qualified staff Good quality paediatric care delivery depends on team work involving a significant number of experienced, trained staff. During this assessment attention was focused on: Findings: the availability of experienced medical personnel; accessibility of doctors and nurses; adequate staffing on all shifts. Less than a quarter (23%) of hospitals (both public and NGO) met the standard for the availability of qualified paediatric staff. In the Iringa and Tabora regions, half of the hospitals had qualified paediatric staff; Shinyanga, Coast and Dodoma regions had only one hospital each. Over half of the hospitals (51%) partially met the standard, whereas in Lindi, Mtwara, Kigoma and Kagera no hospitals partially met the standard. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 27

Figure 12: Number of hospitals with children cared for by qualified staff, by region 4.2.5 Case management Items assessed were: Findings: triage staff able to implement decisions according to the guidelines when the emergency ward became busy. This was a Category A (critical) requirement; skills for managing common emergency conditions; whether treatment started promptly. Overall triage performance was very poor. More than 80% of hospitals were conducting triage that was not based on the guidelines as most staff lacked the required skills, even for common emergency conditions. Sixty seven hospitals were assessed; two were dropped due to lack of data. One Government hospital (1.5%) in the Tukuyu Mbeya region met the standards while 59 (88%) of hospitals did not. 4.2.5.1 Cough or difficulty breathing Poor assessment of a child with a cough or difficulty breathing could easily lead staff to miss a diagnosis of pneumonia, which may cause death. The following Category A (critical) variables were assessed: use of appropriate RCM guidelines for the assessment and classification of pneumonia and wheezing; Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 28

use of oxygen; appropriate antibiotics; appropriate treatment of tuberculosis; and correctly performed chest X ray; adequate performance of the standards significantly improves the quality of care in this area. Findings: A total of seven variables were scored and one item children in need of bronchodilators was dropped because only a few hospitals had this information, which in itself is an indicator of problems with care. Only one hospital (1.4%) had a score of 0; i.e. standard not met due to inappropriate use of RCM. Forty (71%) hospitals partially met the standard and seven (10.7%) met the standard. Hospitals that had missing variable values were excluded from the analysis. They were Biharamulo, Rubya and Kagera in the Kagera region, and Ludewa Hospital in Iringa region. Figure 13: Cough or management of difficult breathing, by region 4.2.5.2 Diarrhoea Similar methods for assessment were applied to obtain information on diarrhoea and five observations were made, two of these being children with severe malnutrition. When not managed properly, diarrhoea can lead to dehydration and cause death. In order to assess the proper management of diarrhoea the following variables were evaluated in the following areas: 1. Assessment of dehydration: was a correct assessment of dehydration carried out (IMCI) (this was a Category A critical requirement); Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 29

was a correct assessment of dehydration in children with severe malnutrition carried out; were infants properly assessed for dysentery, persistent diarrhoea and severe malnutrition. 2. Management according to rehydration plan: was there a correct treatment plan based on the assessment of dehydration (Category A) and fluid therapy; were dehydration and shock appropriately assessed and managed; was there an appropriate treatment plan for each patient; were signs of dehydration monitored during rehydration. 3. Use of antibiotics and zinc: Findings: were antibiotics, particularly for dysentery and cholera, used appropriately; use of zinc for patients with acute and persistent diarrhoea; feeding during diarrhoea (continued breast milk); no use of anti diarrhoea medicines. Six variables were scored and overall performance was very poor as Category A standards were not met in most hospitals. Only 3 (4.3%) hospitals met the standard; 17( 24.%) hospitals partially met the standard; and the majority 45 (5.2%) of the hospitals did not meet the standard. Four hospitals, (Kilolo, Biharamlo, Mpwapwa and Mvumi) had missing variable values for this area of the assessment. Figure 14: Management of diarrhoea, by region Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 30

4.2.5.3 Management of fever Children presenting with fever have differential diagnoses and a thorough assessment is recommended. The checklist for assessment of management of fever had four components; differential diagnosis of fever, and diagnosis and management of severe malaria, meningitis and measles. The variables scored were: Findings: differential diagnosis of fever was properly carried out if children admitted with fever had a differential diagnosis for possible and likely conditions considered. Appropriate investigations were undertaken to establish a diagnosis; if diagnosis of severe malaria was confirmed by laboratory investigations and if its treatment and associated complications (hypoglycaemia/convulsions) were managed as per guidelines (Category A critical requirement); meningitis was ruled out if a lumbar puncture was performed, antibiotics were given immediately and proper monitoring was begun (Category A); measles was ruled out after a proper assessment for complications, whether Vitamin A was given and patients were adequately fed; and whether appropriate investigations were taken. Overall, performance was poor with the majority 3 (52.2%) of hospitals not meeting the standard, as all Category A variables were missed. Nineteen (27.5%) hospitals partially met the standard, and four (5.8%) hospitals met the standard. Nine (13.0%) hospitals, of which (Biharamulo, Kagera, Mugana, Murgwa, Nyakahanga and Rubiya) were in the Kagera region, Ludewa in the Iringa region, Bukombe in the Shinyanga region and Nansio in the Mwanza region, had missing variable values so their scores could not be calculated. 4.2.5.4 Management of severe malnutrition Children with severe malnutrition are often admitted to hospital and death rates during treatment (30 50%) are high. This is as a result of poor treatment practices, lack of clinical guidelines and an inadequate number of qualified staff. Seven variables were scored: assessment of Category A (critical) nutritional status in all children by taking anthropometric measurement (length/height, weight); correct interpretation of anthropometric measurements (Category A requirement); Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 31

appropriate treatment of infections; eye care and immunization; management of severe anaemia and cardiac failure; proper psychosocial stimulation; checking for hypoglycaemia and hypothermia; feeding and distribution of micronutrients and electrolytes. Findings: Overall performance of all hospitals was very poor as with other common clinical conditions. Thirteen hospitals (18.8%) failed to improve the nutritional status of patients. In most of the hospitals length/height or weight were measured but the results could not be interpreted due to lack of clinical guidelines and interpretation charts. Checking for hypoglycaemia was impossible in most hospitals (>85%) due to lack of glucometers and strips. Proper monitoring was not done due to lack of monitoring charts and SOPs. Feeding was poor as calorie requirements were not met and there was a lack of ingredients for micronutrients. With such findings in 0 (87%) of hospitals the standard was not met, in 4 (5.8%) hospitals it was partially met and in 2 (2.9%) the standard was met. Three (4.3%) hospitals (Biharamulo and Rubya in the Kagera region and Nansio in Mwanza) had missing variable values (see Table 14). 4.2.2. HIV/AIDS management HIV/AIDS management covers several areas but this section of the assessment mainly focuses on appropriate counselling and HIV testing, treatment and integration of care and treatment. From this section, three tasks were scored as follows: 4.2.2.1 HIV testing and counselling when appropriate that focused on whether: all sick children assessed for possible HIV and AIDS, including early diagnosis; clinical signs of possible HIV infection identified according to the IMCI algorithm; HIV pre and post counselling provided to parents/care takers and family. 4.2.2.2 Appropriate treatment of HIV and related conditions opportunistic infections were diagnosed and appropriately treated; prophylaxis of PCP given to all HIV infected children; immunization given according to expanded programmes for immunization; regular clinical follow up provided, and palliative and home care. 4.2.2.3 Integration of HIV care and treatment into general hospital services the assessment mainly looked at if there was continuum of care between the different levels of care, and at the existence of links between paediatric services. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 32

Findings: In comparison to the other clinical conditions, HIV/AIDS overall performance was high. Only the Tabora and Lindi regions had poor performance, with Dodoma and Mwanza regions the top performers. Of the 9 hospitals, 21 (30.4%) hospitals met the standard, 29 (42.0%) hospitals partially met and only 18 (2.1%) hospitals did not meet the standard. One (1.5%) had missing variables and was not included in the analysis. Figure 15: Hospital performance scores, by region 4.2.2. Supportive care The section covers topics which were applicable to children with different diagnoses and looked at patient s nutritional needs and promotion of breastfeeding. The nutritional needs of patients were scored using the following variables: nutritional needs of all patients satisfied, according to age and ability to feed; children less than six months old exclusively breastfed at least 10 times in 24 hours; appropriate complementary feeding offered according to IMCI guidelines; children two years and older offered meals three times a day and nutritious snacks twice a day; nasogastric tube feeding is provided for seriously ill children unable to be fed orally, and IV glucose is not used as a source of calories for more than 24 hours; the Category A requirement was for breastfed infants to continue to receive breast milk up to two years of age. Baseline Survey on Quality of Paediatric Care in Tanzania -- Page 33