Our findings demonstrate a substantial decrease in deaths overall in Thailand from 1958 through 1986, followed by an increase beginning in 1987 and then a leveling-off beginning in the early 2000s. All-cause mortality rate trends were similar for males and females, but the rate was consistently higher for males. The gender gap in all-cause mortality rates among persons 25–44 years became more pronounced in recent years, mostly because of fatal traffic accidents and HIV/AIDS-related deaths among men, a group that was more affected by HIV/AIDS in the first phase of the epidemic (12,13). Age-specific mortality rates for children 0–4 years of age showed the greatest decline; the development of and equitable access to maternal and child health care services and vaccination may have contributed to this decline (14,15). The all-cause mortality rate for children 0–4 years of age declined over the study period; however, after the introduction of the Expanded Program on Immunization (EPI), the rate declined 28% (from 715 to 514 deaths /100,000 population). The extensive geographic coverage of primary health care services contributed significantly to maternal and child health outcomes (16).

From 1958 through the mid-1990s, the infectious disease mortality rate in Thailand declined substantially, largely because of declines in malaria, TB, pneumonia, and gastrointestinal infection. Several factors contributed to this trend, including general improvements in sanitation, improved access to medical care (a result of health infrastructure expansions at the district level), and financial risk protection (16), and the introduction of routine childhood vaccination through EPI, which was officially launched in 1977 (14). Since 1987, coverage with TB, diphtheria-tetanus-pertussis, and tetanus toxoid vaccines has been 96%, 75%, and 60%, respectively (17). Deaths related to vaccine-preventable infectious disease declined sharply in association with >90% EPI coverage in the 1990s (18). In the United States, studies have shown a decline in infectious disease–related deaths during the twentieth century (5,6).

In the late 1990s, the decreasing trend for the infectious disease–related deaths reversed. Disease categories that contributed most to this reversal were HIV/AIDS, TB, and pneumonia; all of which had sharply elevated mortality rates during 1997–2003 and decreasing rates in 2004. HIV/AIDS emerged in Thailand in the mid-1980s and spread rapidly with devastating effects (19). In 1999 in Thailand, HIV/AIDS had become the leading cause of death in men 25–44 years of age, resulting in a widening gap in the mortality rate between men and women (8,13). Co-infection with TB and HIV is common; thus; the rising number of TB-related deaths during 1995–2003 coincided with the explosive epidemic of HIV infection and gradually led to the emergence of multidrug-resistant TB (20,21). Moreover, the reported incidences of pneumonia and pneumonia-related deaths had been increasing in Thailand since the mid-1980s (22). It is likely that HIV/AIDS contributed to the pneumonia-related mortality rate. However, this contribution was not readily apparent until later because HIV/AIDS was not a reportable cause of death until 1994 and because stigma was likely a barrier to reporting in the early years of the HIV/AIDs epidemic (23). Furthermore, if the code for opportunistic infectious diseases was used for deaths caused by HIV/AIDS, the number of deaths from HIV/AIDs may have been underestimated (10,24). In persons with HIV/AIDS in Southeast Asia, the common opportunistic infections were TB, cryptococcosis, and Pneumocystis carinii and Pennicillium marneffei fungal infections, all of which can cause pneumonia (25). The observed increase in pneumonia-related deaths in very elderly persons may also reflect the misclassification of the cause of death because pneumonia is often the immediate, rather than underlying, cause of death. A study to verify cause of death data found that ≈31% of deaths coded as being caused by pneumonia should have been classified as being caused by cerebrovascular disease, chronic obstructive pulmonary disease, diabetes, or genitourinary disease (10).

We found that deaths from HIV/AIDS and TB declined during 2004–2009; this decline may account for the concurrent leveling-off of infectious disease–related mortality. Thailand implemented a national AIDS program in 1991 and a national antiretroviral (ARV) treatment program in 2000. The ARV treatment program was designed to increase access to health care and treatment, but it was not until 2004 that the universal ARV treatment program was fully implemented, providing ARV treatment for all eligible patients under the National Access to ARVs for People Living with HIV/AIDS program. As the program scaled up, universal ARV contributed significantly to the reduction in AIDS-related deaths (26–28). Two parallel prevention programs, vertical transmission prevention and condom promotion, contributed to decreasing the incidence of HIV infection and changed the epidemic in Thailand from one that was generalized to one that was concentrated in certain subpopulations, particularly injection-drug users, homosexual men, and youth. TB-related deaths were also reduced through early initiation of ARV treatment program for persons with HIV, implementation of DOTS (directly observed treatment, short course), and appropriate antimicrobial drug regimens for TB treatment (29).

DOTS was implemented nationwide in Thailand in 2001, and since then, the country has achieved the international goal for detecting >70% of the estimated cases of infectious TB (i.e., cases in persons with a TB-positive sputum smear); however, Thailand has not met the international goal for successfully treating >85% of the detected cases (30). The main issues relating to TB control were the fragmented delivery of services; limited capacity of stakeholders and limited coordination between stakeholders; weak referral linkages between hospitals and health centers; high TB/HIV co-infection rates and limited access to ARV treatment, especially among poor persons and those with less education; unsupervised treatment with high default rates; and widespread unregulated use of second-line antimicrobial drugs, which could lead to an outbreak of multidrug-resistant TB and extensively drug-resistant TB (31). Certain infectious diseases (e.g., TB) are re-emerging, and emerging HIV/AIDS epidemics in certain subpopulations have considerable implications for the Thai population. Continued monitoring and evaluation of the effect of interventions on disease incidence and mortality rates are critical if the global goal of curing 85% of TB cases is to be achieved.

We reviewed mortality rates during 1958–2009 in Thailand by using a standardized infectious disease classification scheme. Three possible artifacts in year-to-year fluctuations in the mortality rate over the study period should be considered. The first artifact concerns changes that were made to Thailand’s data recording system during the study period. Four versions of the ICD systems were used, and ICD code changes can lead to substantial changes in long-term trends in cause-specific mortality rates (32). Our results show an increasing trend in deaths from pneumonia and TB since 1994, when the switch was made from ICD-9 to ICD-10 coding; we have not accounted for the differences in diagnostic trends concurrent with the changes in coding. When a US National Center for Health Statistics comparability ratio was applied to the US mortality rates for influenza and pneumonia, it appeared that the declines mainly resulted from the introduction of the ICD-10 coding system (33). We did not apply comparability ratios in our analysis because such ratios were not developed with the data provided by the Thai MOPH.

The second possible artifact is that deaths sharply declined from 1996 to 1997, and this decline was followed by a sharp increase from 1998 to 1999. We cannot rule out that such sudden changes may have resulted from changes in the process for reporting deaths, which was implemented in 1996. In the new death certification system, each death was entered into the computer database in the Civil Registration Database at the Bureau of Registration Administration of the MOI and then transferred to the vital registration database at the MOPH.

The third possible artifact is the cause of death coding errors. The coding of polio deaths since 1997, is an example of such errors. The Polio Eradication Campaign in Thailand was started in 1990, and the last polio case was reported in April 1997 (34). Cases of and deaths from acute flaccid paralysis are aggressively investigated, making it unlikely that a single death could be missed. However, despite the lack of any reported cases of polio since 1997, a total of 274 polio-related deaths were coded during 1997–2009. In general, the quality of death statistics in Thailand is considered poor because many death registrations are incomplete and a large proportion have poorly defined causes of death (11,35). Furthermore, during the 1960s and 1970s, only ≈60% of deaths were registered. That percentage increased to 76% in the mid-1980s and to 95% in the mid-1990s. The highest proportion of unregistered deaths was for infants; this was a result of death occurring, in many cases, before the birth was registered (4,36). The proportion of in-hospital deaths increased from 20% during the 1980s to 43% in 2009 (37). The validity of the overall cause-of death statistics during that time is questionable because many out-of-hospital deaths were coded by persons not medically qualified to determine the cause of death. All of these factors may have influenced the study findings.

This study has some possible limitations. The analysis was based on death attributed to the underlying cause of death as it was reported on death certificates and published in the Report of Public Health Statistics. Because death records list only one cause of death, we knew only the underlying cause of death and, thus, may have underestimated the effect of infectious diseases as a contributing cause of death. As a result, this study may underestimate the role of infectious diseases on mortality rates. In addition, the aggregate nature of our data prevented additional exploration of the specific cause of death by age group. Our estimate of the extent of infectious disease is conservative, focusing exclusively on deaths. This focus reflects only part of the effects from disease because infectious disease may result in substantial illness or disability, or both, without causing death. Future analyses of other dimensions of the effects of infectious diseases, such as their effect on the economy, the number of hospitalizations, and the number of life-years lost because of disability, will provide information to inform policy-making decisions.

The implementation of a malaria control program and new and effective antimicrobial drugs for treating TB contributed considerably to the reduction in communicable disease–related illness and deaths in the second half of the twentieth century (1,38–40). The emergence of HIV/AIDS and the increase in TB- and pneumonia-related deaths in the late twentieth century dramatically interrupted the epidemiologic transition in Thailand. A universal ARV treatment program has rapidly scaled up and resulted in a decrease in the number of deaths from HIV/AIDS after 2003; however, this decline remains unstable. Recent trends emphasize the dynamic process of infectious diseases and highlight the need for sustained resources and efforts to combat their emergence or re-emergence. These data also highlight the importance of addressing health disparities between men and women. Reliable, relevant, and timely mortality data are crucial to guide effective policy responses to protect and promote population health.