Abstract
Cause of death (COD) data is important as it reflects the health problems of a population. The data can be used to assess and monitor the performance of a health system and develop plans for public health work at the national level. Although the country's death reporting system is covered by the civil registration system, it was found that the quality of COD data reported in the civil registry is still poor, especially of natural deaths outside health facilities which depends on information that the deceased’s relative give to the registrar. A Model development for the improvement of data quality on cause of death: A pilot study on deaths outside health facilities in 12 health regions was developed with a primary objective to improve the quality of COD data for deaths outside health facilities according to the policy of the Office of the Permanent Secretary (OPS), Ministry of Public Health (MOPH) with financial support from the Health Systems Research Institute (HSRI). Secondary objectives were 1) to develop a pilot model for COD data quality improvement for deaths outside health facilities, 2) to develop tools and systems for regular review of COD data by community health personnel, and 3) to expand the operational results of improving the quality of COD data for deaths outside health facilities. The study was a research and development study, by developing and piloting a model for improving the quality of COD data for deaths outside health facilities. It consisted of 2 parts: 1) deaths with medical history within 1 year prior to death; medical records (MR) are reviewed by the hospital’s medical coders to ascertain the correct COD in the data collecting program and 2) deaths without medical history during the same period; a community public health officer conducts a verbal autopsy (VA) via an online tool. The VA tool was adapted from the 2016 WHO VA to be short and appropriate for the Thai context for interviews by community public health officer in hospitals and Tambon (sub-district) health promotion hospitals (THPHs). A Provincial Public Health Office (PHO) was responsible for the operations in the province, both monitoring performance and reporting to their Chief Officer. The study was carried out in 12 pilot provinces, one from each of the 12 health regions across the country, namely Chiang Rai, Sukhothai, Nakhon Sawan, Ang Thong, Suphan Buri, Chanthaburi, Maha Sarakham, Udon Thani, Buriram, Ubon Ratchathani, Surat Thani and Songkhla. Death data from the Bureau of Registration Administration (BORA) which was sent to the Strategy and Planning Division, OPS, MOPH was used in conjunction with medical information in the Health Data Center (HDC) database from the Information and Communication Technology Center, OPS, MOPH. The 29,410 deaths between January and May 2022 for natural deaths outside health facilities were allocated to the pilot provinces for review. The study operations included development of tools and systems, training, site preparation, implementation, field visits for reflection, and result summarization, and was carried out between October 2021 and February 2023. Quantitative analysis evaluated the performance according to the pilot model, i.e. percentage of goals achieved. Comparative analysis of COD was performed using COD from death certificates and COD obtained from the pilot study model, i.e. accuracy of COD and decreases in percentage of ill-defined cause. Qualitative analysis using information from meetings with relevant parties, in-depth interviews with executives, and focus group discussions with community health personnel including success factors, problems, obstacles, and solutions, to summarize results and make recommendations. Results Data for natural deaths outside health facilities in the 12 pilot provinces between January and May 2022 included 29,410 deaths. Of these, 28,629 deaths, or 97.3%, were successfully synced and had a medical history within one year before death and/or no medical history but an identified primary care unit. Of the synced cases, 43.8% were deaths outside health facilities with a medical history one year prior to death. There were 781 death cases with no identifiable primary care unit because of a lack of medical history in the HDC database and a lack of details in house registration addresses corresponding to responsible primary care units, such as unspecified municipalities but instead specifying roads or alleys, etc. As of January 3, 2023, data showed that pilot provinces had successful operations for up to 74.1% (21,202) of deaths. A total of 90.3 % of deaths with medical history were reviewed by hospital medical coders, for deaths without a medical history, 61.4% of deaths were followed up with an interview of the deceased’s relatives by a community public health officer. Analysis of COD from death certificates and data from the pilot model implementation found that: 1) In the comparison of COD between death certificates and medical records (MR) reviewed by the hospital medical coders. in 11,318 deaths and 2) In the comparison of COD between death certificates and verbal autopsy conducted by community public health officers. in 9,884 deaths. Factors for success, problems, challenges and solutions. and Policy recommendations.
