Vital Surveillances: Trends and Spatial Pattern Analysis of Typhoid and Paratyphoid Fever Incidence — Yunnan Province, China, 1989–2022

Data on TPF for 1989–2004 were obtained from the Compendium of Infectious Disease Epidemics in Yunnan Province. Reported TPF cases for 2005–2022 were obtained from the infectious disease surveillance system of the CISDCP (3). Additionally, reported TPF outbreaks for 2005–2022 were obtained from the emergent public health event information management system of the CISDCP. Outbreak definitions followed the National Public Health Emergency Information Reporting and Management Specification issued by the Ministry of Health of the People’s Republic of China in 2005 (4). Demographic data were derived from the Yunnan Statistical Yearbook (1989–2022). Administrative division codes and geographical coordinates were acquired from the National Catalogue Service for Geographic Information (https://www.webmap.cn/). The crude incidence rate (per 100,000 population) was calculated as the number of annual TPF cases divided by the total annual population.

Joinpoint regression models were employed to identify incidence trends from 1989 to 2022 using Joinpoint software (version 4.9.1.0; National Cancer Institute, Bethesda, US) (5). The number of joinpoints, joinpoint locations, and P values were determined using Monte Carlo permutation tests. The Bayesian information criterion was used to select the best-fitting model. To explore temporal trends, the APC and AAPC in reported TPF cases and their 95% confidence intervals (CIs) were calculated. An increasing or decreasing trend indicates a statistically significant trend slope (two-sided P<0.05). A stable trend indicates a non-significant APC (two-sided P≥0.05), representing stable incidence or sporadic case reporting (6).

Spatial autocorrelation analysis, using GeoDa 1.18.0.0, explored the spatial correlation strength of TPF. Details on the spatial autocorrelation analysis have been previously published (3). Briefly, global autocorrelation, using Global Moran’s I statistics, analyzed the clustering characteristics of the research objects across the entire region. Anselin’s Local Moran’s I (LISA) test statistics were used for spatial autocorrelation analysis. LISA analyzed the specific cluster types and regions; LISA cluster maps showed four cluster modes: H-H, L-L, L-H, H-L, and not significant. The H-H and L-L regions represent spatial clustering, while the L-H and H-L regions were outliers (3,7).

From 1989 to 2022, Yunnan Province reported 206,066 TPF cases, with an incidence of 32.80 per 100,000 in 1989 and 2.29 per 100,000 in 2022. The average annual incidence was 13.98 per 100,000. A total of 508 deaths were reported, resulting in a case fatality rate of 2.5 per 1,000. Joinpoint regression analysis revealed an overall decreasing trend in TPF incidence, with an AAPC of −6.78% (P<0.05) (Figure 1).

Figure 1. 

Joinpoint regression showing trends in the overall incidence of typhoid and paratyphoid fever in Yunnan Province, China, 1989–2022.

From 2005 to 2022, a total of 75,747 TPF cases were reported in Yunnan Province, including 43,767 laboratory-confirmed cases and 31,980 clinically diagnosed cases. Of these, 48,452 had typhoid fever, with incidence decreasing from 14.44 per 100,000 population in 2005 to 1.88 per 100,000 in 2022. The remaining 27,295 cases had paratyphoid fever, with incidence decreasing from 8.24 per 100,000 population in 2005 to 0.41 per 100,000 in 2022. Additionally, joinpoint regression analysis revealed an overall decreasing trend in both typhoid and paratyphoid fever incidence from 2005 to 2022, with average annual percent changes (AAPCs) of –12.51% and –16.79%, respectively (all P<0.05) (Figure 2).

Figure 2. 

Joinpoint regression showing trends in the overall incidence of typhoid fever and paratyphoid fever in Yunnan Province, China, 2005–2022.

Abbreviation: APC=annual percent change.

From 2000 to 2022, Yunnan Province reported 112,160 TPF cases, with an incidence rate of 20.13 per 100,000 in females and 19.58 per 100,000 in males. Incidence was elevated in the 0–44 age group, peaking in the 25–34 age group (Table 1). From 2005 to 2022, the primary occupations of TPF cases were farmers, students, and children, comprising 44.01%, 19.08%, and 11.35% of cases, respectively. TPF cases were consistently reported from January to December each year, demonstrating clear seasonality. Peak incidence occurred in July and August (Supplementary Figure S1).

From 1989 to 2022, TPF cases were reported in 8 autonomous prefectures and 8 cities in Yunnan Province. In terms of average annual incidence, the top 5 prefectures (cities) were Xishuangbanna Dai Autonomous Prefecture (37.50 per 100,000), Dehong Dai-Jingpo Autonomous Prefecture (28.34 per 100,000), Yuxi City (26.88 per 100,000), Nujiang Lisu Autonomous Prefecture (22.39 per 100,000), and Honghe Hani and Yi Autonomous Prefecture (21.29 per 100,000). Joinpoint regression analysis revealed a statistically significant decreasing trend (all P<0.05) in the reported incidence of 13 prefectures (cities) from 1989 to 2022, except for Yuxi City (P=0.288), Diqing Tibetan Autonomous Prefecture (P=0.332), and Nujiang Lisu Autonomous Prefecture (P=0.468) (Table 2). Furthermore, there was a positive spatial correlation and significant spatial clustering distribution of TPF incidence in all county-level cities/counties/municipal districts of Yunnan Province from 2005 to 2010 (Moran’s I=0.291, P=0.001), 2011 to 2016 (Moran’s I=0.269, P=0.001), 2017 to 2022 (Moran’s I=0.241, P=0.001), and 2005 to 2022 (Moran’s I=0.315, P=0.001). Supplementary Table S1 presents the statistically significant H-H spatial clusters of TPF incidence in a total of 13 county-level cities/counties/municipal districts of 5 prefectures (cities) in Yunnan Province from 2005 to 2022 and the statistically significant L-L spatial clusters of TPF incidence in a total of 22 county-level cities/counties/municipal districts of 8 prefectures (cities) in Yunnan Province from 2005 to 2022.

From 2005 to 2022, Yunnan Province reported 24 TPF outbreaks (12 typhoid fever and 12 paratyphoid fever), with a median duration of 21 days (Supplementary Table S2). These outbreaks involved 1,273 cases, an exposed population of 203,519, and an incidence rate of approximately 625.49 per 100,000. Occurring in 14 counties (districts) across 7 prefectures (cities) in Yunnan Province, the outbreaks primarily affected rural areas (17 outbreaks) and schools (6 outbreaks).

The overall TPF incidence in Yunnan Province, China, shows a significant decrease, potentially attributable to national disease prevention and control policies, public health service development, the inclusion of rural water and latrine improvements in disease prevention and control agencies’ national code of practice (8), immunization, and improved sanitation and hygiene awareness (9). Despite annual declines in TPF incidence rates both nationally and in China, Yunnan Province remains the highest-ranking provincial-level administrative division (PLAD) for these diseases (2). The emergence of TPF as a significant public health issue in Yunnan Province highlights the critical need for effective epidemic control measures. Successfully managing the TPF epidemic in Yunnan Province is pivotal in diminishing the overall incidence of these diseases across China. Indeed, several possible reasons may explain the highest TPF incidence in Yunnan Province. First, Yunnan Province experiences peak TPF incidence during summer due to high temperatures and rainfall, creating ideal conditions for disease transmission (9-10). Second, abundant karst landforms in Yunnan Province increase the vulnerability of underground water sources to pathogenic bacterial contamination, amplifying the risk of TPF epidemics (2,9). Third, the epidemic’s cause may also stem from differences in dietary and water hygiene practices among populations in Yunnan Province’s multiethnic areas (11).

From 1989 to 2022, the top five average annual incidences of TPF were observed in Xishuangbanna Dai Autonomous Prefecture, Dehong Dai-Jingpo Autonomous Prefecture, Yuxi City, Nujiang Lisu Autonomous Prefecture, and Honghe Hani and Yi Autonomous Prefecture. The four prefectures, excluding Yuxi City, are border prefectures with a high concentration of ethnic minorities. The elevated TPF incidence in these areas could be linked to prolonged case accumulation, heightened exposure rates, and changes in dietary and drinking practices among ethnic minorities (11). Additionally, the high TPF incidence between border county-level cities/counties/municipal districts in these prefectures (cities), such as Kunming City, Yuxi City, Honghe Hani and Yi Autonomous Prefecture, and Wenshan Zhuang and Miao Autonomous Prefecture, may be attributed to similar risk factors and provides a hypothesis for cross-regional transmission (12). The high incidence among farmers, students, and children is consistent with the findings of a national study (12) and may be linked to poor living conditions, increased outdoor exposure, and inadequate dietary and hygiene practices.

However, this study has limitations. First, data on TPF cases were acquired from the CISDCP infectious disease surveillance system via passive surveillance, potentially introducing reporting bias (3). Second, differences in testing, diagnostic, and reporting capabilities of hospitals at different levels in various regions lead to bias in identifying and reporting TPF. Third, the spatial autocorrelation analysis scale selection depends on the researcher’s subjective judgment and does not consider the temporal characteristics of clustering; false positives are inevitable, so these results should be interpreted cautiously. Finally, this study did not include driving factors (e.g., pathogen resistance) and facilitating factors (e.g., meteorology) (10) that may influence TPF incidence; therefore, the causes of TPF incidence could not be analyzed. In conclusion, while the reported TPF incidence in Yunnan Province has decreased notably, it remains high, with noticeable spatial clustering in certain prefectures (cities).