CONTENTS. 1. Introduction. 2. Lessons learned from MDR and MDSR Who is this guideline for? Goals and objectives 7. 5.

Transcription

1

2

3

4 CONTENTS PART I: THE BASIS AND BACKGROUND 1. Introduction 3 2. Lessons learned from MDR and MDSR 3 PART II: THE TECHNICAL GUIDELINE for CDSR 3. Who is this guideline for? 7 4. Goals and objectives 7 5. The scope The overall scope of CDSR the three components Scope with regard to place of death Scope with regard to age of child Confidentiality and anonymity Assessing preventability of child deaths Assessing substandard care contributing to death Considering child protection concerns 9 6. Operating principles Have clarity on the objective of CDSR Create and enhance community awareness Generate commitment, confidence and collaboration Build on what exists Nurture and sustain collective learning Organizational considerations The levels for conducting child death audit Expertise for conducting CDSR Linkage with MDSR The responsibilities and terms of reference of each level 12 i

5 8. The specific inputs for the process for child death review (CDR) Notification of child deaths in hospitals and in community The number of deaths to be notified, triangulation of data The review process when to initiate The tools/formats used for the review The methods of review The number of deaths to be reviewed at each level Frequency of review at each level Determination of cause of death Identifying the factors leading to the death, and assessin 22 preventability 9. From review to response Timing of responses Responses at the different levels Guiding principles for response Surveillance of child death Types of surveillance Data entry and quality check for completeness Data aggregation, analysis and interpretation Analytic plan and indicators Trend analysis, and more complex analysis Reporting, feedback and dissemination Who should receive feedback and reports? Method and channels of dissemination Periodicity of reports 29 ii

6 PART III: THE IMPLEMENTATION PLAN 12. The initiation (pilot project) phase Why? Where? When? How? Translating and pre-testing the forms Training Monitoring of progress Phased implementation and scaling up Advocacy and gaining support What is advocacy? How to carry out advocacy, the successful factors Target groups for advocacy Monitoring and Evaluation Monitoring Supervision Evaluation The indicators Resources needed for implementation 41 ANNEX 1 : Lessons learned from MDR/MDSR 43 ANNEX 2 : CDSR Form 1 Notification form for child deaths 47 and investigation of the three delays ANNEX 3 : CDSR Form 2 - Verbal autopsy form for child deaths 53 ANNEX 4 : CDSR Form 3 Summary report of TMO 67 ANNEX 5 : CDSR Form 4 Summary report of DMO/District CD Review Team 70 ANNEX 6 : Checklist for supervision and quality check by TMO on midwife 73 ANNEX 7 : Logical Framework for monitoring and evaluation 75 iii

7 ACKNOWLEDGEMENT Under the overall guidance and leadership of the Ministry of Health (MoH) especially Dr Myint Myint Than, Director, Child Health Division, Department of Public Health, this technical guideline was developed during the period from March to May MoH, first of all, would like to extend its deep appreciation to members of the Lead Child Health Working Group for their valuable inputs in pulling together this Child Death Surveillance and response (CDSR) technical guideline. Special thanks goes to Dr NarimahAwin, international consultant, for her wonderful facilitationand valuable insights which she brought to the table. High gratitude also goes to UNICEF s MNCH team for their unlimited technical support throughout the process of CDSR Guideline development. Great appreciation is also extended to the other divisions and departments under the Ministry of Health, the Central Statistical Organization,State/Region/Township Health Department, WHO, UNFPA, INGOs, local NGOs and 3 MDG Fund for their active participation and great contribution. Last but not least, sincere thanks also go to the representatives of Basic Health Staff particularly Midwives as without their perspective and contribution, development of the realistic and meaningful technical guideline will not be accomplished in such a smooth manner. i v

8

9 2 Technical Guideline for Child Death Surveillance and Response

10 1. INTRODUCTION The death of a child 1, a young life full of promise and potential, is a tragedy. A large number of child (under-five) deaths are preventable, by interventions that are efficacious, effective and affordable. An important component of any elimination strategy is surveillance. A surveillance system counts and tracks the numbers and trend of child deaths, but also helps to understand the underlying factors that contribute to the deaths and how they can be corrected and the death prevented. For this to happen, a Child Death Surveillance and Response (CDSR) should be put in place along with guidelines for the development and implementation of the CDSR. Besides surveillance, a central component of CDSR is the elicitation of the cause of death and the factors that contributed to the death, which is achieved by conducting a death review or audit, on which the third component, response is founded. A maternal death review (MDR) system already exists, and efforts have begun to change this to a maternal death surveillance and response (MDSR) as recommended by the Commission on Information and Accountability (COIA) of the UN Global Strategy for Women s and Children s Health (GSWCH). In addition, Myanmar has updated its National Child Health Strategic Plan (NCHSP), and the plan for is ready for roll-out. This strategic plan makes special mention of the need for developing a CDSR as one of several efforts to improve the information system on child health, especially to better understand the circumstances surrounding a child death 2. LESSONS LEARNED FROM MDR AND MDSR In Myanmar, a community-based maternal death audit began as the first pilot in five townships of Sagaing Region in 2005, and it was intended to include perinatal (and later replaced by neonatal) death review, but this was not implemented. CDSR guidelines have been developed with lessons learned from the MDR/MDSR experience, by asking two questions (1) What were the tools used for neonatal death audit; will these be appropriate for use in CDSR (2) What lessons can CDSR learn from MDSR?. The answers to these are shown in ANNEX 1 1 A child is variously defined in different contexts and for different purposes; the Convention of the Rights of the Child defines it as below the age of 18. For legal purpose, different countries may use different age limits. In this guideline, a child is defined as from birth to 5 years old (0 to 60 months) of life, and child mortality is often referred to as under-five mortality. In the context of death review this guideline includes still births Technical Guideline for Child Death Surveillance and Response 3

11 4 Technical Guideline for Child Death Surveillance and Response

12

13 6 Technical Guideline for Child Death Surveillance and Response

14 3. WHO IS THIS GUIDELINE FOR? The Technical Guideline is for health care providers, health professionals, health care planners and managers, and policy makers working in the area of child health who strive to improve child survival by improving access to and quality of services; improving information on child death, both quantitatively (measuring accurately the burden of child death) as well as qualitatively (by identifying and understanding the factors and circumstances, notably the preventable and avoidable in order to better formulate strategies to prevent child deaths. This guideline will also be useful for nongovernment and civil society organizations (NGOs and CSOs) including faith-based and religious organizations involved in child welfare and health. Stakeholders with the ability to drive change for improving child survival should be involved in all aspects and processes of the CDSR to ensure that child deaths are notified and reviewed, responses are implemented and surveillance is in place. 4. GOALS AND OBJECTIVES Goal To contribute to a more comprehensive and reliable information system in child health and survival that will effectively leads to a reduction in child mortality in Myanmar. General objective To provide information on child deaths, through the conduct of a child death review that identifies the causes and contributing factors to the deaths, followed by making the appropriate responses, and carrying out continuous surveillance on child deaths Specific objectives 1. To receive notifications on as many child deaths as possible, preferably on ALL child deaths 2. To collect accurate data on number of child deaths by relevant variables 3. To collect information on the causes of child deaths and the contributing factors, to understand how and why the deaths occurred 4. To identify and assess the contributing factors that were avoidable or preventable 5. To make evidence-based recommendations for response to decrease child death 6. To contribute to an effective system of surveillance of child deaths Technical Guideline for Child Death Surveillance and Response 7

15 5. THE SCOPE 5.1. The overall scope of CDSR There are three components: 1) Death Review, 2) Surveillance and 3) Response. The Death Review component involves the analysis of information of causes and contributing factors of child deaths and monitoring the trend at various levels of the health system The Surveillance component requires strengthening of existing surveillance systems in child health, and these only need to be strengthened to meet the purpose of CDSR. Surveillance in general exist in three main forms, and each will be applied to child health and survival (i) routine surveillance through the HMIS, which can cover the whole country or from selected sites as sentinel surveillance (ii) periodic surveys such as DHS, MICS and verbal autopsy for causes of child deaths (for which a tool is available) 2 (iii) ad-hoc surveys and studies based on an identified need. CDSR itself can contribute to surveillance of child death The Response component follows all reviews with recommendations for actions. The response must be robust, documented and monitored 5.2. Scope of CDSR with regard to place of death Child deaths can happen either in facility/hospitals or outside the facility (home, school and other locations in the community). In Myanmar a large proportion of child deaths are in the community. To meet the objective of CDSR which is to understand the circumstances/factors contributing to the death so that in future such deaths can be prevented, deaths in both settings must be audited. The CDSR shall include child deaths in facility (but only in public facilities) and in the community. At this early stage, deaths in private facilities shall not be subjected to the audit process, but the incident should be reported to allow for surveillance to be carried out (only for tracking the number over time), and currently the HMIS captures and counts the number of child deaths from private facilities Scope of CDSR with regard to age For the scope of CDSR, any stillbirth beyond the gestational age of 22 weeks and any death of a child below five years will be included. This includes neonates, postneonates, infants and older children from one to under five years. Therefore it is proposed that 2 Verbal Autopsy Standards: The 2012 WHO Verbal Autopsy Instrument Release Candidate 1 8 Technical Guideline for Child Death Surveillance and Response

16 The CDSR shall consist of audit of deaths in two age cohorts stillbirths and neonate (0-28 days) and child (29 days to under 5 years). The notification form will be common for the two age cohorts The verbal autopsy form (for a detailed investigation of some deaths that the TMO or DMO or review team deem necessary) shall also be common for the two age cohorts, and this shall be the form already used in the 2013 study on cause of death There shall be only one committee or review team to conduct these reviews. The MDR committee that already exists shall be expanded in mandate, membership and terms of reference to include CDSR 5.4. Confidentiality and anonymity in the death review process The CDSR aims to uncover causes and circumstances surrounding a child death in order to identify preventable or avoidable factors so that similar deaths in future do not occur. There is no punitive intent, and therefore there is no need to keep the names of care providers anonymous and confidential Assessing preventability of child death Most child deaths in developing countries are preventable by simple effective and affordable interventions. The optimistic outlook of prevention has led to the global target of elimination of preventable child deaths. In the context of CDSR, the concept of preventability or avoidability is used in the review/audit of individual deaths, and the assessment of factors that led to the death whether they were preventable. A child death can be classified as avoidable if the death might have been avoided by a change in patient behavior, provider/institutional practices, and health care system policies. It is recognized that the determination of avoidability does not follow rigid criteria and it is often open to interpretation. Preventability often overlaps with substandard care 5.6. Assessing substandard care contributing to death Often, the preventable factor uncovered by the review is a form of substandard care for example, a child s life could have been saved if antibiotics were available to treat the pneumonia he was suffering from. 5.7 Considering child protection concerns Determining child abuse or neglect as a cause of death is outside the scope of the CDSR, but any suspected cases should be reported to the police for investigation. Technical Guideline for Child Death Surveillance and Response 9

17 6. OPERATING PRINCIPLES The operating principles must support all three essential components of the CDSR: 1) Death Review, 2) Surveillance and 3) Response. (See 5.1 Scope of the CDSR) Have clarity on the objective of CDSR A child death review or audit is a comprehensive review of child deaths to better understand how and why the child died, and to use the findings to take action to prevent other deaths. It is not meant to be a punitive exercise, and the findings are not to be used for litigation or proof of malpractice Create and enhance awareness The health staff especially those involved in the processes of CDSR need to be aware of the importance of CDSR, and what benefits it can bring. Increased awareness among health providers can lead to changes in knowledge and practice in caring for a sick child; among health managers and policy makers, it will lead to changes in policies and reallocation of resources. The community should also be aware of the importance of CDSR and understand that it will reduce child deaths, and that the community can have a role especially in reducing or eliminating delays in care seeking behavior Generate commitment, confidence and collaboration Health staff, after being made aware of the importance of CDSR must be committed to ensure its success, and they must be given knowledge and skills (and fear of punishment is dispelled) for them to gain confidence. The principle of anonymity and confidentiality (presented as scope of CDR in the preceding section) to some extent can allay this fear because the health staff (and the patient) is not identified in the review process. Collaboration among the levels of health care, and among disciplines within the health care is crucial to ensure a good effective CDSR 6.4. Building on what exists There is already a child death notification form, which serves as a review or audit tool. Forms were developed for neonatal death review which were not rolled out. The HMIS is a robust system which can provide many of the required information. There has been a verbal autopsy on cause of child deaths in a survey. All these tools have been used to develop the CDSR Nurture and sustain collective learning The review process involving a multi-disciplinary range of experts provides a good opportunity for learning from each other. It is important for the team to promote shared responsibility and teamwork, and foster collective learning for action at different levels, including lessons for policy, clinical practice and community action. 10 Technical Guideline for Child Death Surveillance and Response

18 7. ORGANISATIONAL CONSIDERATIONS An organisational structure, especially death review teams with clear terms of reference and responsibilities is critical in CDSR. In CDSR there is need for the involvement of the district level. The main functions of identification and notification of deaths; review of the death; analyse and recommend; and response will involve the following structure (individuals, team, levels of care), each with their expertise, responsibilities and terms of reference The levels for conducting child death review The different levels for CDSR especially the conduct of the review/audit will be essentially the same as that for MDSR, which starts at the operational area where the midwife is stationed, to township to district and region/state and then to national level. For CDSR, the district level is important because the need for determining cause of death as accurately as possible requires a paediatrician. There is a paediatrician (and some other specialists such as ObGy) in the district hospitals. Therefore the child death review team 3 consisting of the various relevant expertise shall be established at the district level. Above the district level, the Region/State Review team shall be established but its functions will be that of analysis and collation. At the national level, the National Review Team conducts even less review, and more in-depth analysis to see overall trends. These responsibilities are described in the Section 8.4 below. It is necessary to appoint a CDSR Coordinator (who shall be a member of the Review Team) at each level to facilitate communication and receipt of forms, records and reports. The coordinator shoud be deputy regional/state health director Expertise for conducting CDSR A child death review requires specific expertise. At the operational level where the death is notified, the midwife is considered an expert in this function of receiving and recording the notification, and also for the filling of the death notification/ review forms. The TMO and his/her team members shall have the requisite knowledge and skills to go through the review forms submitted by the midwives in the township, and skills to collate the information, carry out basic analysis and write a summary report. The same applies to the DMO, but at the district level, there shall be a Child Death Review Team whose members represent experts in paediatrics, neonatology, obstetrics, perinatology or maternal-feto medicine (if available), public health, nursing, midwifery and any other as needed. An important expertise at this level is to check the cause of death assigned by the midwife and attested by the TMO, to either confirm or to amend the cause of death 3 Various terms are used for the group of experts who conduct a death review/audit Committee, Panel, Team. In this Technical Guideline, as agreed by the CHLWG, the term used is TEAM Technical Guideline for Child Death Surveillance and Response 11

19 Expertise in the review teams at region/state and national level shall consist of mainly public health, epidemiology, health programme management, paediatrics, neonatology, perinatology, obstetrics and any other as and when needed. At central level, there should be a representative from the Department of Medical Research 7.3. Linkage with MDSR There shall be one review team to review maternal and child deaths at the various levels but there has not been MDR committee at district level. The terms of reference of the existing MDR committee at all levels must be reviewed. The members shall be the same (except there is a need to expand this to other subspecialists such as neonatologist and maternal-feto specialists). The activities being planned for MDSR including training shall take into account, wherever relevant, the needs and plans for CDSR, and vice versa The responsibilities and terms of reference of the different levels The responsibilities and functions of the various levels are described here. However these functions require specific inputs (including tools) that are described in Section 9 1. The midwife is the focal point for all child deaths reported from the community a. Auxilliary Midwife (AMW), Community Health Worker (CHW), village leaders/community volunteers etc will inform the midwife of any child death in the community. In contexts where there is no midwife in the village and it is possible for the AMW/CHW, the AMW/CHW shall fill in the CDSR Form 1. These will be shared with the midwives on a monthly basis. The AMW/CHW shall have the required education and training to do this. b. The midwife will fill in in Form 201 (requirement of CSO); and the prescribed CDSR forms for notification of child death, as well as for investigation of the three delays (CDSR Form 1 see Section 9.4) either in consultation over telephone or in person with a direct contact of the child (who could be a health worker who provided treatment or parent/caregiver who accompanied the child during the illness) or receive it from the CHW/ AMW. c. Midwife transfers all notification forms to the Township Medical Officer (TMO) on a monthly basis during the monthly CME at the township while retaining a copy d. If the midwife has heavy workload and requires assistance, the Public Health Supervisor Grade II (PHS II) can fill up the required notification form 12 Technical Guideline for Child Death Surveillance and Response

20 2. The Township Medical Officer (TMO) and team at township health office a. Reviews all the CDSR Form 1 (notification/review forms and the three delay investigation) submitted by the midwives b. Analyses all notification forms for accuracy; in case of inaccuracies, TMO verifies with the concerned midwife and rectifies death notification form c. If he/she feels it necessary, TMO shall conduct, or direct the HA or LHV at the concerned health unit to conduct, a verbal autopsy using the prescribed forms (CDSR Form 2 - see Section 9.4) d. Prepares the quarterly summary report (CDSR Form 3 - see Section 9.4) which he submits to the DMO e. Reports through a reporting system to Region/State Health Department and Child Health Department on a monthly basis this entire process will be automated with introduction of mobile technology, while retaining a copy, whenever it becomes available 3. The District Medical Officer (DMO) a. On a quarterly basis after receipt of notification forms from every township, reviews the cause of death as per need, based on discussion with TMO b. Identifies 5 top causes of death and randomly picks up 3 notification forms per cause of death during that quarter, from each of the two age cohorts - hence selecting 15 notification forms from each (total 30 forms). If there are fewer deaths than the required number then all deaths will be selected for review c. May also identify any other cause of death to be reviewed, such as if it is epidemiologically important, for example any illness that is consistently increasing over a period of time, or any illness that is not acceptable to be a cause of death d. Convenes the Child Death Review (CDR) Team on a quarterly basis to conduct the death review The District Child Death Review team comprises of the DMO, all TMOs in the district, and medical specialists from relevant disciplines. This Team has the following TORs: a. Reviews all the 30 selected notification forms (and any other identified for review) in detail, ascertains the immediate cause of death and the underlying Technical Guideline for Child Death Surveillance and Response 13

21 cause of death. In case further clarifications are needed, the team calls the concerned midwife to investigate further b. Identifies gaps and bottlenecks in the system that have contributed to these deaths, identifies actions or intervene (immediate and long term) that can prevent similar deaths in future, indicating who is responsible for implementing these actions c. Shares the outcome of the audit with all townships, Region/State Health Departments and the Child Health Department MOH, within one week of the District CDR team meeting, through a summary report (CDSR Form 4 - see Section 9.4) d. If needed the team may decide to trigger a verbal autopsy in which case, the a member of the district team along with the TMO and midwife will travel to do a verbal autopsy (VA); some indications for VA are (but not limiting to) causes of death that do not come down after a period time of more than one year, in spite of successful implementation of interventions, unusual causes of death that cannot be investigated by midwife or TMO; causes of death that are epidemiologically important as identified above, death due to possible criminal acts, etc. 4 e. During every meeting the team reviews the actions taken by both the District and the Township based on the previous quarter and those that have not been acted upon or have been inadequately acted upon; these need to be identified as priorities in this report in addition to new action points as identified by current quarter review 4. Region/State level at this level, a CDR review team is also established but its functions will be less of review and more of analysis a. Analyses data based on the monthly reporting format (on a quarterly or semester basis) to profile the townships, identify epidemiological trends that need immediate interventions, and share the analysis and actions identified based on a standard reporting format with Child Health Department, every semester/ 6months b. Analyses the report from the District CDR team on a semester basis, track actions, check quality of reports, identify areas for improvement and provide feedback to District CDR teams and TMOs based on finding; report the outcome of analysis and actions taken, to Child Health Department 4 All health staff should be sensitized to the possibility of child protection concerns, such as child abuse or neglect, which should be reported to the police for investigation. 14 Technical Guideline for Child Death Surveillance and Response

22 c. Based on the findings of the analysis, conduct appropriate responses- e.g. improve training programme, increase monitoring and mentoring, trigger VA studies, modify or introduce programme interventions etc 5. National level (Department of Child Health. MOH) a. Analyse data based on the monthly reporting format, on a semester or annual basis, to profile the townships, identify epidemiological trends that need immediate attention b. Analyse the report from the region/state on a semester or annual basis, track actions taken, check quality of the reports, identify areas for improvement and provide feedback to region/state c. Based on the findings of the analysis, conduct appropriate responses e.g. improve training programme, increase monitoring and mentoring, trigger VA and other disease surveillance, modify or introduce programme interventions d. Produce an annual report of CDR which profiles the cause of death, key interventions/actions identified, status of implementation of these interventions, trend in that year, compared to previous year, priorities for the next year, etc e. Present findings of the CDSR to the CHLWG at suitable intervals (either 6 monthly or annually) The functions are as follows. Membership and chairmanship of the Review Team at district level the chair of the Review Team is to be based on co-chairmanship with the Department of Public Health and Department of Medical Service to chair each review meeting or chair on a rotation/alternate basis. Members of the team are selected on area of expertise, which can be from both Public Health and Medical Service. Procedures for deaths that occur in hospitals are reviewed (or not reviewed) in the hospital (See Section 8.1). Technical Guideline for Child Death Surveillance and Response 15

23

24 8. THE SPECIFIC INPUTS FOR THE PROCESS OF CHILD DEATH REVIEW (CDR) Some of the specific inputs and actions (particularly the tools/forms) may differ between the two age cohorts of deaths to be reviewed (i) neonatal deaths (ii) deaths in children from 29 days to below one year: 8.1. Notification of a child death A child death can occur in a health facility or outside a facility in the community. Information flows from the community starting with the midwife, who fills up the prescribed forms currently on paper or receives them from the trained CHW/AMW (See Section 8.4) Reporting and review of deaths outside hospital / in the community - The midwife will be made aware of a child death in the community either from a notification from the family (for the family near to where the midwife is stationed) or from her periodic field/home visits, or from community volunteers (AMW, CHW, village head, community volunteers, etc). It is therefore critical that as many deaths are reported in this manner, and it is suspected that some (even many) child deaths are unreported, especially stillbirths and deaths of very young children. All efforts must be made to encourage the community to report death of a child at every age including stillbirths. The midwife fills in the required form (CDSR 1 see section 9.4). If there is no midwife in the village and the CHW/AMW is educated and trained in the use of the forms, they will fill this form. Reporting and review of hospital death - One can reasonably assume that all deaths (including child deaths) in a hospital are reviewed/audited by the hospital as a clinical audit as part of quality assurance. In Myanmar while this is encouraged, it is only carried out by some hospitals. In addition, in big hospitals (with 150 beds or more), there is already a perinatal and neonatal database. There are existing tools for these reviews, therefore these death reviews will not use the CDSR forms as described in this guideline. For hospitals that carry out death audits, it is necessary that they are reported to the TMO (for station and township hospitals), DMO (for district hospitals) and Region/ State health department for Regional hospital. These levels will ensure that these hospitals deaths which have been reviewed by the hospitals are counted in the death records/statistics. If a death in a hospital is reported but has not been reviewed, the respective level will see that the review is carried out by the relevant staff in the hospital using the CDSR forms in this guideline. Technical Guideline for Child Death Surveillance and Response 17

25 8.2. The number of deaths to be reported/notified, triangulation of data The intent of any death review system is to gather as complete as possible the information related to the deaths, so that it is adequate for the formulation of strategies and interventions to prevent future deaths. Use of community volunteers should be encouraged to report child deaths to the midwife. It is important to avoid duplicating notification of the same child death. Triangulation of data between sources using personal identifiers can be helpful to ensuring each death is reported only once. For instance, a facility and community may both report the same death. The midwife and community volunteers who inform the midwife (AMW, CHW, village heads and others), must ensure that deaths are not reported and counted more than once 8.3. The review process - when to initiate It is hypothesized that a verbal autopsy conducted too many days after the event will be less reliable. The midwife should try to conduct the review (CDSR Form1) as soon as possible, within 7 days after the death notification. This initial review, while it does not amount to a full in-depth verbal autopsy, does involve an interview of a family member of the deceased child or a health care provider who cared for the child before death. More in depth interview (verbal autopsy) may need to be conducted as deemed necessary by either the TMO or DMO, based on the forms submitted The tools/forms used for the death review The forms used consist of (a) Notification and brief review form (in which is also incorporated the investigation for any delay in seeking care) (b) Verbal autopsy form, in the event that a death reviewed as part of the notification needs further review/ investigation by detailed verbal autopsy and (c) Summary forms/reports by TMO and DMO a. Notification, brief review and investigating for delay in seeking care: Upon a child death being made known, the midwife fills up Form 201 which is required by the CSO and is filled for all deaths. For the CDSR, the midwife fills a common death notification form, which is common for the two age cohorts (CDSR 1) ANNEX 2. Part 1 of this form is the notification of death with causes indicated. Part 2 of the form seeks to elicit some basic information on any delay in the seeking of care for the child using the Three-Delay 5 model. It is recommended that in circumstances where the workload burden needs to be lessened for the 5 The first delay is failure in recognising the need to seek care; the second delay is although there is recognition to seek care, there are barriers to do so such as geographical, cultural or financial barriers; the third delay is when after having sought care, the care provided is of poor quality 18 Technical Guideline for Child Death Surveillance and Response

26 midwife, this task of filling CDSR Form 1 can be carried out by the PHS II. Currently the volunteers (AMW and CHW) are only responsible to inform the midwife of the death. In areas where the AMW/CHW are educated and are trained to fill in this form, they can be do this and submit to the midwife on a monthly basis. b. More detailed review by verbal autopsy: The TMO or DMO may request after the review, that a verbal autopsy on some of these deaths be conducted, using the CDSR 2 Form (ANNEX 3). This verbal autopsy can be carried by any staff from the township as directed by the TMO, or by the Health Assistant (HA) in charge of the RHC or Lady Health Visitor (LHV) where the death was notified from. This verbal autopsy shall not apply to stillbirths. The form has 10 sections with separate sections for cause of neonatal (Section 8) and post-neonatal deaths (Section 9) and a section (Section 6) for deaths due to accidents. It also has several questions in Section 10 (reports and returns) if a death certificate was issued, and it requires the interviewer to see the death certificate, and record the causes of death (immediate and underlying) and contributing factors stated in the certificate c. Summary forms/reports: At the township level, the TMO after receiving all these forms from all midwives in the township, shall fill up the Summary Report A (CDSR Form 3) as in ANNEX 4. The TMO also sends all notification/review forms to the DMO, not for the District to conduct a review but to confirm or ascertain the cause of death in each from, because the review to be conducted at this level is on a sample/selection of forms to be selected based on cause. At the district level, the DMO (and District CD Review Team) after receiving the notification forms reviews the cause of death. The DMO, in consultation with the paediatrician, sees all the forms, confirms or amends the cause of death, compiles the forms by cause, takes the top 5 leading cause in each age cohort, randomly takes 3 forms from the top 5 causes, (thus deriving 30 forms) and then the District Review Team conducts death review on these 30 deaths. If there are fewer than the required number the maximum available should be reviewed. The District CD Review team shall also prepare a summary report, based on the summary reports submitted by the TMOs - CDSR Form 4 (ANNEX 5). Technical Guideline for Child Death Surveillance and Response 19

27

28 8.5. Methods of child death review The methods of review to be used shall be mainly facility-based review (of medical records) and a verbal autopsy (for deaths in the community) The number of deaths to be reviewed at each level The number (and frequency see 8.7. below) of any death review depend on the number of cases identified and the resources available to collect the necessary data. Unlike maternal deaths, the number of child deaths is much larger, as many as two hundred-fold more. Child death audit shall be conducted on a sample of deaths, to be determined by a reasonable and practical method. This sample shall be based on cause of death, even though we recognize the low accuracy of cause of death especially for deaths in the community. However surveillance in terms of numbers reported and tracking this number over time and by location shall be done for all under-5 deaths It has to be borne in mind that a death review/audit is a time consuming process at every level, right from the midwife who initiates the process who is likely to be required to fill a long format. The midwife and member/experts of the review committee carry out this function as an additional task over and above their normal tasks. The number of deaths to be reviewed at the various levels are: The midwife reviews all child deaths that are reported to her using the notification form, (CDSR Form 1) which incorporates the three-delay investigation The TMO reviews all the forms sent by all the midwives in the township; it then sends all of these to the District Medical Officer (DMO). If TMO finds it necessary, a verbal autopsy is conducted using CDSR Form 2. At the District level, selection is to be made on the number of deaths to be reviewed. It is possible that the district can receive as many as 60 death reports (with 5 or 6 townships making up a district). As has been described earlier, the number to be reviewed by the District Team is arrived at by first picking the top 5 causes of death from each of the two cohorts (neonatal deaths and deaths in children aged 29 days to 5 years), and then taking randomly 3 forms from each top cause in each cohort, thus deriving a total of 30 forms selected for review. At Region/State and National level, only the summary reports from the District are received, unless the review team requests for specific information from some of the notifications. These levels conduct an in-depth analysis, detecting any epidemiological trends, and assessing the responses carried out at the district and township levels Technical Guideline for Child Death Surveillance and Response 21

29 8.7. Frequency of review, and structure for review at each level The frequency of review will be quarterly as described earlier. The midwife sends in the audit forms to the township every month; the TMO and DMO conduct quarterly reviews. All TMOs will participate at the quarterly review meeting at district level; and if the TMO is not able to attend, he/she shall be represented by a staff from the township office. The Region/State conducts reviews every six months and at National level, the review is done annually Determination of cause of death Determining and assigning the cause of death requires a good level of clinical knowledge and acumen. Additionally, it also uses ICD coding and classification. It is acknowledged that the level and coverage of knowledge of this is relatively poor in Myanmar. This will require additional training. For neonatal deaths, the causes are also quite clear and covers a narrow range of possibilities preterm births (often with respiratory distress syndrome), birth asphyxia, sepsis, congenital malformations. In the case of death of older infants (post neonate) and children below 5, the cause of death covers a wider range of possibilities. Therefore, CDSR Form 1 for notification requires the midwife to assign the cause of death in terms of sign/symptom (fever, cough, etc) or a clinical diagnosis (tetanus, measles, malaria a case definition protocol for common conditions. Once these forms are submitted by the midwife to the township level, this cause of death may or may not be reassessed. It is when the reports reach the district level that a paediatrician is available to review/reassess the cause or causes assigned to each death, and a final cause of death is determined. The TMO submits all notification/review forms to the District level, where each form is assessed for the cause of death assigned by the midwife and TMO, and where this cause is either confirmed or amended based on the best evidence available to the District Review Team, as has been described earlier 8.9. Identifying the factors leading to the death, and assessing preventability Knowing the clinical cause of a child death does not provide all the information needed for understanding why the child died. The next step is to look for the events before the death, especially to identify any preventable factor that could have averted the death; and if there was any possibility of substandard care being given. From the findings from the three delay investigation from (CDSR Form 1) an assessment can be made if there was a delay in failure for the family to recognise the need to seek care (Delay 1), failure /inability to do seek care (Delay 2), after having sought care and reaching the point of care, was the quality of care adequate (Delay 3). 22 Technical Guideline for Child Death Surveillance and Response

30 For deaths in hospitals, in addition, the medical records of the management of the child will be used. For this the TMO writes a narrative report for each notification/ audit form he receives, on the preventability of the deaths, as the last item in the summary report (CDSR Form 3). These shall be reviewed by the District CD Review Team, which shall fill up CDSR Form 4 - in which it writes a concise narrative on the preventability of deaths in the district in that quarter The CDR shall identify preventable factors, and shall not be used to punish staff, for reasons mentioned earlier in Section 6. Preventability and substandard care can overlap, for example - shortage of antibiotics or other life-saving supply/commodity, a staff who came in too late to attend to the child, staff not skilled in neonatal resuscitation, etc. 9. FROM REVIEW TO RESPONSE One of the weak points in many death review systems is the review-to-action link. Following the review, recommendations for Review will be documented with responsibilities for action assigned to specific staff and with time lines. Responses need to be appropriately and optimally timed, and at the various levels of the health system. There are some general guiding principles related to response Timing of response Response following a review may be immediate or periodic Immediate response - Findings from reviews of nearly every child death can lead to immediate actions to prevent similar deaths from occurring. This may be easier in the case of deaths at health facilities but is true for deaths both at home and facility. A child death review can identify gaps in areas that should be addressed quickly both in health facilities and communities. Child deaths in health facilities often point out needed improvements in quality of care such as inadequate coverage of emergency services by skilled providers; lack of essential medications or supplies; need to improve knowledge or skills of providers in the management of child illnesses; or need to improve overall services. Deaths that take place in communities can also identify some actions that can be implemented quickly. There is no need to wait for aggregated data to begin implementing actions to prevent child deaths. Periodic response - Monthly, quarterly or semi-annual reviews (depending on numbers) of aggregated findings should take place at larger health facilities and at the district level where there is a District CD Review Team These periodic reviews may begin to show a pattern of particular problems that are contributing to child deaths, or particular geographic areas where child deaths are occurring Technical Guideline for Child Death Surveillance and Response 23

31 in greater numbers. Such findings should result in a more comprehensive approach to addressing the problem across multiple facilities or across multiple communities. Where areas at greater risk are identified, discussion with the involved communities should be a priority to identify solutions. Annually - every health facility should summarize its child mortality findings annually. In larger health facilities where multiple deaths may have occurred, the findings should contribute to continuous quality improvement plans. At the township and district level, findings from the analysis of aggregated data and the aggregated recommendations from the death reviews are incorporated in a district report. Actions at the district level may include health-system strengthening and staff retention, resource mobilization, increasing community and institutional awareness of maternal mortality, fostering community-facility partnerships and building alliances with the private sector, and advocacy activities Responses at the different levels Responses can be at a health facility at the various levels, or in the community. Community level: Responses may include actions in the community. Findings from the community may point to the need for the development of health promotion and education programs as well as possible changes in community service provision; changing home practices or the attitudes of health care providers; or improved infrastructure such as roads, bridges, and communication technology. Some findings, like the latter, may require a longer time period to plan, implement and obtain the necessary government support. Information from health facilities may point to the need for changes in clinical practice or modification of service provision Health facility/hospital: A child death in a health facility such as a hospital should be considered an unacceptable event that needs a review. Each death, if properly reviewed, should identify systemic problems that contributed and can be corrected. These may include (1) staffing issues whether there are sufficient staff to meet the demands for quality child health care including essential newborn care (2) knowledge and skills including all those who are involved in providing care or supportive services; and (3) deficiencies related to infrastructure, medications, equipment and other supplies that may have led to inadequate management of complications At higher management levels: At township, district, region/state, there will be appropriate responses depend on the findings of the review. At the national level, analysis is conducted from all aggregated data that contributes to a national child health 24 Technical Guideline for Child Death Surveillance and Response

32 plan.. Actions may include allocating required resources to the most affected areas and populations. Actions at the national level may also include changing or updating national policies, laws or guidelines. Responses taken in the previous quarter shall be an agenda item in each CD Review Meeting at district level, and that shall be one of the functions in the list of responsibilities of the District Public Health Officer 9.3. Guiding principles for responses Some guiding principles for response are Start with avoidable factors identified during the death review process Make sure response is evidence-based, however remember that not all problems identified have evidence-based solutions, particularly those related to family, community, transportation issues and access to care. Ideally, actions that are not based on known evidence will be evaluated to ensure they are having the expected effect. Prioritise the responses It is likely that many problems will be identified; not all can be tackled simultaneously so it is important to prioritize them. Some characteristics should be considered when prioritizing problems and their solutions. One important factor is prevalence¾how common is a problem? Resolving common problems may have a greater impact than resolving unusual problems. Another factor is the feasibility of implementing the intervention. Is it technologically and financially possible? Are there sufficient human resources? What are the costs? Finally, what is the potential impact of the intervention? If it were successfully implemented, how many children could be reached and how many lives saved? Decide how to monitor progress, effectiveness, impact - Remember, CDSR unlike CDR requires that responses are not only robust, but also monitored and documented Integrate recommendations within annual child health strategic plans Finally, it is useful to bear in mind that recommendations should be specific and link with avoidable factors. 10. SURVEILLANCE OF CHILD DEATHS For this third component of CDSR, we need to recollect the types of surveillance, and then proceed to the handling and management of data and information, which is the central feature of surveillance Technical Guideline for Child Death Surveillance and Response 25

33 10.1. Types of surveillance In Section 6.1, it is stated that while the Surveillance component may imply a new input, in almost all countries there are existing surveillance systems for child health, and these only need to be strengthened for meeting the purpose of CDSR. The CDSR may use any of the three main methods of surveillance (i) Routine surveillance through the HMIS, which can cover the whole country (as in Myanmar), and if necessary supplemented by sentinel surveillance (ii) Periodic surveys such as DHS, MICS and verbal autopsy for causes of child death, as have been carried out using the WHO tool (iii) Ad-hoc surveys and studies based on an identified need. CDSR itself is a form of surveillance of child death which contributes information to both the number (and trend) as well as the factors contributing to child mortality Data entry, quality and completeness In preparation for analyses, a clear framework for data transmission, aggregation, processing, and storage needs to be defined. In the initial stages of implementation of CDSR, the data collected are relatively simple and not voluminous. However the check for quality and completeness are important so that analysis is based on valid information. The CDSR coordinators at the various levels will check the information found in the notification/review forms Data aggregation, analysis and interpretation Data analysis and interpretation of results are critical components of any surveillance system. The initial data analysis should be done at the level closest to the community with the appropriate analytic skills. This is the township level and more robustly at district level where there is a Child Death Review Team with various expertise among the members. Through CDSR, all facilities will know their facility specific number of child deaths, be able to calculate indicators for their facility, and report on the causes of death that occur in their facility. Each child death must trigger the question why did it happen? and, when appropriate actions are available, immediate responses must be set in motion. The aim of aggregated data analysis is to identify causes of death, identify factors contributing to the deaths, assess the emerging data patterns, and prioritize the most important health problems At the district level, the CDSR coordinator will maintain a data-base and checks for completeness and inconsistencies between data items. The review team will be notified of any problems, if necessary, including inconsistencies or inadequate reporting of certain items. The review team will also be informed of differences encountered in the number of entries and asked to verify the counts or to determine 26 Technical Guideline for Child Death Surveillance and Response

34 the nature of the inconsistencies. The database will be utilized for analyses of all the child deaths that have been reviewed. The review team fills in CDSR Form 5 (summary report) and sends it to Region/State When performing CDSR analyses, the following factors are prerequisites: Knowledge of surveillance (sources, mechanisms, data collection instruments, completeness of reporting, abstraction, data entry and validation) Good understanding of the indicators to be calculated and denominator issues Changes over time in case definitions, detection, or data collection Knowledge of the limitations of the data, such as incomplete coverage, poor quality, and changes over time in data processing may also influence the analysis. These prerequisites imply additional resources there is need to move towards computerisation and electronic data management at district and national levels, and to assign a dedicated information (HMIS) personnel for CDSR at national level. At the moment, one focal person should be from child health division of the Department of Public Health Analytic plan and indicators An analytic plan is important to guide the analytic process and identify problems in the health system that may contribute to child deaths, especially those that are amenable to change. The plan should include: the identification of appropriate and feasible indicators prior to data collection; guidelines to calculate rates and proportions and how to display data; how to compare rates with expected or reference values. When the volume of data becomes large, statistical probability methods may be necessary to examine apparent. Interpretation of the findings should focus on aspects that will lead to prevention of death Trend analysis, and more complex analysis On-going surveillance can provide more detailed information about changes over time (temporal trend), as well as over place (spatial trend). Specific analyses can be conducted at district level to identify patterns and trends. These may be used to influence district action and response. At region/state level, it can show which districts are in greatest need, similarly at national level, the regions/states most in need can be detected. More complex analyses may be needed to answer specific questions that arise. These analyses may require approaches beyond what are routinely performed. Time series analyses and analyses using geographical information systems (GIS) are very valuable approaches that should be considered when appropriate resources exist. Technical Guideline for Child Death Surveillance and Response 27

35 11. REPORTING, FEEDBACK AND DISSEMINATION There should always be a feedback of the findings and the recommendations down to the level of the hospital or the community where the information was collected. Government accountability for child health requires the periodic and transparent dissemination of key results, particularly child mortality and the progress made in achieving international goals such as the MDGs Who should receive feedback and reports? Any death review systems must build in a system for the dissemination of findings and recommendations to stakeholders who have direct interests: Individuals and agencies within the formal health system who provide information and are involved in the process of CDSR, from the midwives to the TMO and other staff at township level to, district and state levels, for them to take immediate actions. Senior management staff in the health system need to be made aware of the findings in order to formulate develop strategies and interventions, and strengthen the current system. Policy makers need to be aware of the findings of the CDSR to furnish them with the information and evidence for support and commitment, and to increase their awareness about the magnitude, social effects, and preventability of child mortality. The impact of interventions needs to be shared with those who are involved in implementing these interventions. This awareness and commitment will lead to specific actions including allocation of resources more effectively and efficiently by identifying specific needs. Feedback also enhances accountability for child health, including accountability for getting reliable data for calculation of child mortality rates which calls for efforts toward complete civil registration/ vital statistics. It will guide and prioritize research related to child mortality. Dissemination of findings to relevant government agencies should be pursued, especially the agencies whose mandate covers determinants related to child survival, including those that have a role in reducing the first and second delays (poverty alleviation, transport, etc.) Dissemination to agencies outside the formal health system and government agencies can be considered academic institutions, professional bodies, NGO/ CSOs, the private sector 28 Technical Guideline for Child Death Surveillance and Response

36 Dissemination methods and channels The findings from CDSR can be disseminated using a variety of channels to enable a wide range of people to access it and ensure that the information gets to the right audience, namely those who can act on the recommendations. Methods for dissemination of results can take many forms, for all interested parties (including health staff) some are more appropriate for a certain level, for example At community and first level of facility, townships Regular team meetings, thematic seminars held at facilities, regular Village Health Committee and Rural Health Centre meetings, community meetings, radio programmes, printed reports, training programmes, posters, text messages, video clips At subnational (district and region/state) and national level - Printed reports for policymakers, statistical publications, scientific articles, professional conferences. training programmes, media, press releases, websites, newsletters and bulletins, fact sheets, video clips Periodicity of reports Reports are to be generated by the various levels with reasonable periodicity. The township and district levels are to generate quarterly reports that focus on salient findings, and recommendations for immediate action and suggestions for improvement in the next quarter. These need not be comprehensive reports. For the Region/State level and national, an annual report that highlights the findings of the audit including the responses, as well as what the surveillance system has shown should take place. An epidemiological profile of the child deaths in the region/state, and the country is given in the report Periodic workshops to disseminate and share the findings will be conducted at state/region level, and an annual workshop held as a national workshop Technical Guideline for Child Death Surveillance and Response 29

37 30 Technical Guideline for Child Death Surveillance and Response

38

39 32 Technical Guideline for Child Death Surveillance and Response

40 The implementation of CDSR will necessarily be a slow incremental process that is likely to take several years, starting with a pilot project. 12. THE INITIATION FIRST PHASE Why? Where? When? How? A First Phase will be implemented to see the feasibility of the CDSR processes as described in this technical guideline, and to detect any problem in the implementation, so that changes are made before the system is implemented in a wider area, until nationwide coverage. The First Phase shall be in the townships that are already receiving funding by the 3MDG Component 1 initiative. A total of 36 Townships are in this category. These townships will be selected, recognizing that that some of these townships are not suitable either due to particular challenges such as geographical remoteness or because they are in conflict areas. A training manual for the staff in the pilot townships will be developed, and after the training is conducted, the First Phase will commence. It is reasonable to assume that enough lessons can be learned after 12 months of this first phase of implementation Translating and pre-testing the forms The forms CDSR Form 1 (part 1) and CDSR Form 2 are already in current use and do not require translation into the Myanmar language, and require no pre-testing. Part 2 of CDSR Form 1 on investigation of three delays is new and will need translation before pre-testing. Pretesting can be conducted in a conveniently located township(s), which is expected to need only one day to two days. The summary forms (CDSR 3 and 4) for the TMO and DMO respectively will be reviewed at the national workshop and will not need pretesting; it will need translation into Myanmar Training of staff in pilot townships There is a need for training of the staff in the First Phase area and the following steps carried out: training plan/module developed this is not a complex exercise, because the training needs to only give a brief background of CDSR, and practical sessions on how to use the tools/forms, of which CDSR 1 and CDSR 2 (for the midwives) are already being used, and they need to be familiarised with form CDSR 3; while the TMOs and DMOs need to be familiarized with the Summary Reports A and B. trainers need to be identified Director of Child Health and senior staff of the Department, and selected members of the CHLWG. Technical Guideline for Child Death Surveillance and Response 33

41

42 realized, and before that the scaling up shall occur according to three dimensions (Figure 10.1 taken from the MDSR guideline, but is applicable to CDSR). a. The first (horizontal) dimension is the place where the child death occurs for the initial phase, the CDSR shall capture both facility deaths (but only public facilities) and deaths outside the facility (at home, school, playgrounds, and other sites in the community). It is envisioned that in future, CDSR will be conducted for deaths in private facilities (for the audit and response components because the surveillance component in terms of number of deaths is already in place with the HMIS). b. The second (vertical) dimension is the geographical coverage of the CDSR itself, especially the audit component. This will be better assessed and decided after the completion of the First Phase. c. The third (diagonal) dimension is the extent and depth of the review/audit process. It is proposed that the CDSR in Myanmar at this initial stage of introduction, uses only medical (hospital and other facility) records and the verbal autopsy conducted to complete the two notification/review forms (CDSR Forms 1 and 2) supplemented by CDSR Form 3 which identifies any delay is seeking care. In later phases of implementation, these tools shall be improved or expanded or new tools introduced. Needless to say, the extent and speed of implementation depends on several factors. The enabling factors in MDR/MDSR will apply to CDSR, and these include (1) political will and a clear agenda for child health (2) good collaboration and communication among the different levels of review (3) Knowledge and skills of staff involved in the processes especially of death review (4) proper and complete documentation (5) user-friendly and simple tools and formats, supplied in adequate numbers (6) checklists and supervision, and (7) adequate resources 14. ADVOCACY AND GAINING SUPPORT From the experience with MDSR, advocacy is found to be useful, it is one of the seven thematic areas 6 in the country road map to implement the recommendations of the Commission on Information and Accountability (COIA) under the Global Strategy for Women s and Children s Health (GSWCH) What is advocacy? Advocacy is understood as an organized, deliberate systematic process intended to bring about a positive change This definition has been adopted in the country road 6 The seven thematic areas are 1.Civil Registration and Vital Statistics (CRVS), 2.Monitoring and evaluation, 3. Maternal Death Surveillance and Response (MDSR). 4. Innovation and e-health, 5. Monitoring and tracking of resources, 6. Review process and 7. Advocacy and outreach Technical Guideline for Child Death Surveillance and Response 35

43 map for implementation of the recommendations of COIA. In the Technical Guidelines for MDSR, advocacy is defined as A political process by an individual or group that aims to influence behavior, policy and resource allocation within political economic and social systems and institutions 7 Changes in behavior and clinical practice are often difficult to achieve without widespread promotion and visible support from leading and well respected advocates, professionals and professional organisations How to carry out advocacy for child health and CDSR? Advocacy can happen in many different ways, and choosing the best option depends on what needs to change to improve child survival in Myanmar. Some of the topics that can be used in advocacy exercise in the context of CDSR are: Highlighting the magnitude of the problem, which is unacceptable level of child mortality in Myanmar, and the fact that most of these deaths can be prevented by effective and affordable intervention Highlighting the socio-economic dimensions of child death especially inability to access health care and life-saving interventions Demonstrating patterns and trends of child mortality and the slow progress made towards international goals and targets Exposing bottlenecks to influence change eg. access to drugs Identifying gaps or absent protocol or policies Successful advocacy takes rigorous, in-depth research, careful planning, and clearlydefined practical goals. It needs clear purpose, well-framed arguments and sound communication with audiences Target groups for advocacy for CDSR Four groups should be targeted as priority: Health professionals, for them to be aware of the need for CDSR, and how CDSR can benefit the health status of children Parliamentarians, in order to collect support for CDSR including procurement of government resources The media, which use a variety of channels for a wide audience Civil society organisations (CSOs), which can harness community resources 7 Maternal Death Surveillance and Response Technical Guidance : Information for Action to Prevent maternal Deaths, WHO (2012) 36 Technical Guideline for Child Death Surveillance and Response

44 15. MONITORING AND EVALUATION Monitoring While monitoring is a critical activity in CDSR, the scope of monitoring lies within the three components of CDSR which are death review, surveillance and response, and not for the child health programme. For example, if the service is found to have a shortfall in quality which led to the death, it is not the function of the CDSR M&E system to remedy this this is the function of the child health programme managers, who should respond to findings of CDSR and improve the programme. The M&E of CDSR on the other hand shall detect shortfalls in the CDSR processes themselves, and take remedial actions to improve these processes. In this regard, monitoring of the CDSR system is needed to ensure that the steps in the system are functioning adequately and improving with time, and that information is adequate and timely. Monitoring of the CDSR system should be carried out at all levels Supervision Supervision and quality checks are essential components of monitoring. In the context of CDSR, supervision is necessary and to be formalised for the midwife. The checklist for supervision of the midwife is in ANNEX 6. Supervision and quality checks of the midwife are carried out according to the normal hierarchy in the health system, as follows: In the existing organisational protocol, the midwife must be supervised by the Lady Health Visitor (LHV) or the Health Assistant (HA) in the rural health unit. Supervision for CDSR shall be conducted along with the regular supervision that is already being done; there is an existing supervision checklist. Therefore the supervision checklist for CDSR (ANNEX 6) will need to be incorporated into the existing checklist. The midwife along with her supervisors (LHV and HA) will discuss with the TMO when she submits her monthly forms and reports at the monthly meeting with the TMO. The DMO and CD Review Team at district level discusses with the TMOs in the district who are invited for the quarterly death review meetings. This meeting is an opportunity for the DMO to do quality checks with the TMOs. If deemed necessary, the DMO may make a visit to the townships (perhaps on a rotational basis for one visit per township in a year) to observe for himself/ herself the way the TMO manages the CDSR process, especially with regard to ascertaining cause of death. While a checklist is not necessary, the DMO should write a report of his supervisory visit to the TMOs. In a similar manner, the state/regional supervisor may wish to make a supervisor visit to districts for any particular reason, including to observe the Technical Guideline for Child Death Surveillance and Response 37

45 functioning of the District CD Review team during the audit meeting. It is suggested that one district receives one supervisory visit per year from state/ regional supervisor especially with regard to epidemiological analysis and trends of child death and its profile. While a checklist is not necessary, the supervisor should write a report of his supervisory visit to the DMOs Evaluation In addition to the monitoring indicators that provide a quick snapshot of whether the system is improving, periodically a more detailed evaluation is useful. The main purpose of CDSR is to lead to action to reduce child deaths, therefore if this is not happening the system is not meeting its objectives, and the system should be implemented using resources appropriately. Effectiveness - In the early stages of implementation, evaluation is limited to effectiveness to determine if the recommendations for action have been implemented (inputs, processes or activities), if they are achieving the desired results (outcomes) and if not, where any problems may lie. Exactly how this effectiveness evaluation should be carried out will depend on the particular circumstances in each community, facility, or health care system. It starts with a determination of if and how the specific CDSR findings and recommendations have been implemented. Impact This evaluation will be to assess whether the system has achieved its goal reduction of child deaths and lowering of child mortality. Efficiency - Later, evaluation can extend to efficiency, to see whether the system can function more efficiently. This includes an assessment of its key processes: identification and notification, review, analysis, reporting and response. IT solutions can help reduce inefficiencies, but require trained staff. Ideally the system will be computerized, at all levels in a phased manner. Midwives can be provided with mobile phones to improve speed of notification and review Indicators The indicators will be for effectiveness (processes/output and outcome) and impact; no attempt shall be made in the early stages of implementation to measure and identify indicators for efficiency or cost-effectiveness. This however may become relevant in the later stages especially when there is need to procure resources and justify for them. Effectiveness (process and outcome) indicators: These are considered together in the following table with the assumption that the input/processes/output will lead to changes which are the outcomes. In a relatively small project such as CDSR (unlike projects or programmes with wider scope and activities) it is convenient and reasonable to place inputs, processes/activities and outputs together. 38 Technical Guideline for Child Death Surveillance and Response

46

47

48 16. RESOURCES FOR IMPLEMENTATION Besides the First Phase implementation, which has to be costed separately, the implementation of CDSR will depend on the availability of required resources, which at the very least will be for: Time and the travel needed, with their cost implications in the evaluation of implementation of MDR in by SEARO, some midwives stated that they often had to pay for travel to reach the deceased s house to conduct the verbal autopsy. Printing of tools/formats in adequate numbers - the evaluation of MDR in revealed that the forms were often in short supply and midwives sometimes had to make photocopies at their own cost Supervisory and monitoring visits by supervisory staff Printing of periodical and annual reports Advocacy activities use of media (print or radio/television etc) has cost implications. In the road map of MDSR, briefing sessions are recommended for parliamentarians, media and civil society organisations media Training that may be needed from time to time Convening of meetings, seminars, workshops etc related to CDSR Resources for the development of electronic formats, systems and tools, so communication technology can be used to facilitate timely and comprehensive reporting in the longer term. Computerization at all the levels of review will also facilitate the CDSR, particularly for the review and surveillance components. Technical Guideline for Child Death Surveillance and Response 41

49 42 Technical Guideline for Child Death Surveillance and Response

50 ANNEX 1 Lessons learned from MDR/MDSR experience This ANNEX provides answers to the two questions: (1) What were the tools used for neonatal death audit and will these be appropriate for use in CDSR? (2) What lessons can CDSR learn from MDSR? Technical Guideline for Child Death Surveillance and Response 43

51 ANNEX 1 : Lessons learned from MDR/MDSR experience Q1 : What were the tools developed under the MDR for neonatal death audit, will these be appropriate for use in CDSR? In Myanmar, as mentioned earlier, there was an attempt to introduce perinatal (later on replaced by neonatal) death audit, using the MDR platform. However, this component did not get implemented as planned, and did not progress as MDR did; justifiably because maternal death is given a much higher priority than perinatal/ neonatal death. An attempt was made to introduce neonatal death audit with the following forms (i) Cause of neonatal death. While this detailed form on cause of death can be used as a verbal autopsy tool in CDSR, it is proposed that this will not be used, as it is too detailed and complex for the midwife (ii) Referral form (iii) Form to identify the three delays (iv) Terms and definitions used in verbal investigation of newborn death. For the purpose of CDSR, these forms are not needed except the cause of neonatal death, for which there is an existing form - the verbal autopsy forms developed by the WHO 3, and Myanmar had used this before in 2013 study on cause of child death. Q2 : What lessons can CDSR learn from MDSR? The lessons learned from the MDSR experience include: Maternal death review is beneficial: MDR has shown that the understanding the causes of maternal deaths and the of factors and circumstances surrounding the death provides an excellent opportunity for formulating policies, strategies and actions to prevent several maternal deaths, and reduce maternal mortality Maternal death review is feasible: The experience with MDR has shown that it is feasible to elicit and understand the factors and circumstances that lead the maternal death, in other words to go beyond the numbers or statistics on mortality rate and ratio; so that an opinion can be formed if the death was preventable or avoidable, and what could have been done to prevent the death. There are of course lessons for CDSR to learn from MDSR, and these include Methods of death review: The methods of review are likely to be the same five methods described for MDSR found in the WHO guide Beyond The Numbers, except the near-miss method. Similarly, some relevant information can be obtained from the guide on MDSR especially in the flowcharts for reporting, auditing, responding and conduct of surveillance Getting reliable sources of information: Any death audit system relies heavily on reliable information system and sources of data especially from civil registration and vital statistics (CRVS). This has been found to be a major problem in MDSR. The MDSR guidelines recommend that countries make maternal death a notifiable event, which may not be possible/realistic for 44 Technical Guideline for Child Death Surveillance and Response

52 child deaths. The M&E Technical Strategic Group of the M-HSCC s proposing to the GFATM for funding to strengthen birth and death registration and mortality analysis in Myanmar, which will directly benefit the development of CDSR Difficulty in determining cause of death: Information is particularly lacking in determining cause of death. In MDSR, there is a further inherent difficulty of definition and classification of death, which will not apply to CDSR a death in a woman of reproductive age may be pregnancy-related death, but not necessarily a maternal death, and indeed it is only through a careful audit of pregnancy-related deaths can the maternal deaths be identified 9. In the information on 863 maternal deaths in 2013, the assignment of the cause of death was inaccurate The number of deaths: Child death far outnumbers maternal deaths by as much as twenty times or more. MDSR has called for every maternal death to be notified, counted and preferably all are reviewed. Indeed in some countries where the number is too small to allow lessons to be learned, audits are done for maternal near-miss cases as well. For CDSR, only a sample of deaths can reasonably and practically be reviewed. Selection of variables for the review process: The investigation/review tools for CDSR such as reporting formats will be guided by the example of MDSR including the selection of variables to be studied. In the analysis of 863 maternal deaths in 2013, the quality of the analysis was especially compromised by information on the cause of death and the three-delay model When to initiate review?: It was reported by a few midwives that the 7-day protocol to initiate the verbal autopsy was not appropriate (too soon) for some families, who could not adequately respond to the interview while they were still in a state of grief Staff knowledge, attitude and motivation: It has to be recognized that a death review involving specific actions and tools requires specific knowledge. In addition, attitude can also be a problem. Adequate training has to be conducted. It has been reported in several studies that sometimes staff deliberately not report a maternal death for fear of being punished for possible substandard care that the MDR may reveal. While this may not be the case in CDSR, it is still critical that health staff are informed from the beginning that the audit is 9 A pregnancy related death includes maternal deaths and deaths from incidental causes to which the pregnancy, childbirth and puerperium has not contributed. The maternal deaths (which exclude these incidental deaths) may be direct maternal deaths due to obstetric causes or indirect maternal deaths due to underlying conditions such as heart disease etc Technical Guideline for Child Death Surveillance and Response 45

53 not for punitive purposes. Sometimes, the higher level managers themselves were not convinced of the importance of MDR, and therefore did not give enough support. Staff workload: The MDR is an extra burden on the workload of the staff; hence it is critical to factor this in when developing the system. It has to be borne in mind that a death audit is an extra task/function of the midwives whose existing responsibilities are already heavy to begin with. In Myanmar, some midwives found the 32-page audit form too long and takes time to fill; in addition, the forms are often in short supply Supervision and support: The support and supervision given by senior managers was found to be critical for ensuring that the midwife conducts the audit properly, and eventually for the success of MDR Logistical problem: Besides the shortage of the audit form mentioned above, there were reports of difficulties faced by the midwife in transport to the facility of to the house of the deceased to conduct a verbal autopsy The management and administration: For CDSR, this can be modelled after MDSR such as the committees and review teams, logistics involved, processes such as reporting and dissemination (indeed this technical guideline recommends that the same review committee be used for both MDSR and CDSR The cycle of actions: Finally, the overall approach to a death review used in MDSR is a cyclical approach. These four cyclical steps are applicable to CDSR (i) identifying deaths, (ii) reviewing the deaths, (iii) analyzing the findings of the review and (iv) response with an ongoing monitoring and evaluation (surveillance) 46 Technical Guideline for Child Death Surveillance and Response

54 ANNEX 2 CDSR Form 1 : Notification Form for Child Death and questionnaire to investigate the three delays This form is filled by the midwife who receives information that a child has died. Part 1 of the form is a simple form (not detailed as in a verbal autopsy) with fixed options on possible cause of death based on signs and symptoms. If needed (midwife not able to do so), this form can be filled by the PHS II Part 2 of the form his form is for the midwife (or PHS II) to investigate if there was any delay in seeking care based on the three-delay model Technical Guideline for Child Death Surveillance and Response 47

55

56

57

58 Q1: Were you aware that you had to take the child to a clinic or a hospital for the illness? No stop Yes go to Q2 Q2 : Did you or someone else take the child to the clinic/hospital? No go to Q3 Yes go to Q4 Q3 : Why did you not take the child to the clinic/hospital? (Allow the respondent to explain in his/her own words but be aware of the possible reasons that may need to be probed by some leading questions see LIST 1). Q4: Did you have any difficulty in taking the child for care as soon as you decided that your child needed further care? Not sure stop Yes, I had some difficulty and I was late in taking my child for further care? go to Q5 and 5a No, I had no difficulty and I took my child on time go to Q6 Q5 : What difficulty did you have in taking the child on time to the clinic/hospital? (Allow the respondent to explain in his/her own words but be aware of the possible reasons as in Q3 and the LIST 1) Q5a : After you arrived at the clinic/hospital, were you satisfied with the services provided? Yes Stop No go to Q7 Q6 :Were you satisfied with the services or care given? Yes stop Not sure-----stop No do to Q7 Technical Guideline for Child Death Surveillance and Response 51

59 Q7 : What are the reasons you were not satisfied with the service /care given at the clinic/ hospital? (Allow respondent to answer freely but be aware of the possible reasons see LIST 2).. LIST 1 - Possible reasons for not seeking care or delayed in seeking care (Delay 1 and 2) - patient factors Delay 1 Not aware of need to seek care Did not think the illness was serious and can cause death Wanted to treat the child at home Went to alternative health provider - traditional healer, quack Delay 2 Clinic/hospital too far, cannot get there easily No transport available No money for transport No money to pay for medical care and medicines No one to look after other children at home LIST 2 : Possible reasons for dissatisfaction with the service at point of care (Delay 3) health system factors There was no health staff Clinic/hospital was closed There was no available bed Staff came late to attend to the child Staff told us nothing could be done to save the life of the child Staff did not appear skilled or competent Staff was unfriendly, rude Staff not willing or not able to answer questions We were informed there was no medicine and we had to buy them ourselves The clinic/hospital was dirty/uncomfortable/had no basic amenities 52 Technical Guideline for Child Death Surveillance and Response

60 ANNEX 3 CDSR Form 2 : (Verbal autopsy from for child deaths) This form is to be filled up for any child death notified in which the Township Medical Officer (TMO) or District Medical Officer (DMO) or the District Maternal and Child Death Review Team/Committee deems it necessary to be investigated further. The TMO or DMO or any team member may decide to conduct the autopsy themselves, or may direct the Health Assistant (HA) or the Lady Health Visitor (LHV) at the RHU to conduct the autopsy This form is available in Myanmar language in which it has more details (compared to the English version below) following revisions and editions Technical Guideline for Child Death Surveillance and Response 53

61

62

63

64

65

66

67

68

69

70

71

72

73