Addressing the Unaddressed: Challenges in Geospatial Mapping of Typhoid Fever Cases in Ghana Open Access

Geospatial mapping has been fundamental in the success of public health interventions during disease outbreaks since the 1850s when Dr John Snow first implemented it during London’s cholera outbreak.¹ In recent years, many large-scale projects have successfully utilized geospatial modeling data in resource-limited settings, such as evaluating malaria incidence in Sudan and identifying HIV risk factors in Mozambique.^2,3 However, challenges arise in some resource-limited areas when performing geospatial mapping within smaller geographic areas due to a paucity of accurately updated and detailed maps.

In northeastern Ghana, typhoid fever remains endemic, creating a significant public health issue with high rates of morbidity and mortality. The disease is caused by Salmonella typhi, whose only natural host and reservoir is humans. Key prevention strategies include safe food and water, personal hygiene, and proper sanitation. However, as of 2020, over 5 million people in Ghana continued to rely on surface sources of water, and more than 80% did not have access to improved sanitation and toilet facilities.⁴ This is especially evident in the North East region where only 22% of homes have water standpipes for potable water.⁵ 

Typhoid fever presents with nonspecific signs and symptoms, and the sensitivity of diagnostic testing (such as positive blood culture) remains low. Therefore, in regions with high disease incidence, diagnosis is often made based on clinical criteria including fever, abdominal pain, and diarrhea. Diagnosis is further supported when certain known complications occur such as intestinal perforation near the ileocecal valve.⁶ 

Spurred by an observed uptick in cases of typhoid fever intestinal perforations, we conducted a geospatial mapping project involving patients treated at a regional referral hospital in Ghana’s North East region. The study’s primary goal was to identify potential clusters of cases and any possible sources of contaminated water nearby. The project’s secondary goal was to mitigate further cases via focused public health interventions near identified clusters.

This study aims to address the following research questions:

1. Are there identifiable geographic clusters of typhoid fever intestinal perforations in the North East region of Ghana?

2. What are the challenges and limitations of conducting geospatial analysis in this resource-limited setting?

3. How can the results of this geospatial analysis inform targeted public health interventions?

This retrospective study was conducted at Baptist Medical Centre in Nalerigu, Ghana, located in the North East region, reviewing cases from September 2023 to January 2024. The study was approved by the hospital’s deputy director of nursing services and head of research. The study met criteria for internal review board exemption. Due to the retrospective nature of the study using deidentified data, informed consent was waived.

Due to the lack of reliable blood cultures at the facility, typhoid fever diagnosis was based on clinical criteria, including fever, abdominal pain, and peritonitis, along with intraoperative findings consistent with typhoid perforation. Inclusion criteria consisted of patients who underwent a surgical procedure for suspected bowel perforation due to typhoid fever between September 2023 and January 2024. Exclusion criteria comprised patients with intestinal perforations from other etiologies or patients with incomplete medical records. We reviewed charts of patients who underwent exploratory laparotomy for suspected intestinal perforations during the study period.

We collected basic patient demographics, including age, sex, weight, duration of symptoms, and referral status. The patient’s hometown or village and preferred language were also recorded. Although Ghana has a digital addressing system (known as Ghana Post Global Positioning System or GhanaPostGPS), this information was not available in medical records because the system is not widely used in the region. The data was extracted from electronic medical records (Lightwave Health Information Management System), and the medical records were reviewed by the primary study investigator (K.T.).

Google Maps, which uses the World Geodetic System 1984 geographic coordinate system, was used to locate each patient’s hometown or village. For locations not found on Google Maps, we conducted focused interviews with hospital leadership, community members, and patients to approximate the location of the town or village. These focused interviews were conducted by 2 of the authors (K.T. and P.A.) together to ensure there were minimal to no language barriers. In the North East region of Ghana, several villages have duplicate names and can only be differentiated by the predominant language spoken in each village. The patient’s preferred language listed in the medical record was used to determine the correct village location. We defined a “reasonable degree of certainty” for mapping as consensus among at least 2 local sources familiar with the area’s geography.

Point and heat mapping were performed using Arc Geographic Information System Pro version 3.3.1 by the Environmental Systems Research Institute. We analyzed the spatial distribution of cases and attempted to identify any significant clusters using heat map symbology as a representation of relative density for visualization. Descriptive statistics were used to summarize patient demographics and clinical characteristics.

During the study period, we identified 87 cases of intestinal perforation due to typhoid fever. Three cases were excluded due to missing data. Of the resulting 84 cases, the patient’s hometown or village was easily identified on Google Maps in 38 cases (45%). For the remaining 46 cases, we were able to map 36 locations (43%) with a reasonable degree of certainty based on focused interviews. Data on the patient’s preferred language was available for 79 patients (94%). Ten locations (12%) were unable to be mapped with any method and were therefore not included on the final maps. In total, the patient’s hometown or village was mapped with a reasonable degree of certainty for 74 cases (88%).

Patient demographics revealed that most cases were in pediatric patients, with 85% of patients under the age of 18 years. The average patient age was 13.2 years. There was a slight male predominance, with males accounting for 59.5% of cases. The average duration of symptoms at the time of hospital admission was 7.7 days. Approximately one-third of patients (31%) had been referred to the hospital from another facility. Notably, more than a third of patients were underweight for age at the time of admission (36.9%). Underweight for age was defined as weight-for-age less than the fifth percentile on Centers for Disease Control and Prevention growth curves.

Two maps were created with the obtained map coordinates. The first map is composed of raw point data indicating the patient’s hometown or village for each mappable case of typhoid fever intestinal perforation (Figure 1). However, in this view, it is difficult to appreciate the 34 (46%) overlapping points when more than 1 case occurred in the same village. Therefore, data was also illustrated with heat mapping (Figure 2). The area with highest density is in the region surrounding the study site.

The largest number of cases occurring in the same town was 8; however, this town is a district capital with an estimated population of approximately 11 000 people.⁷ More cases would be expected from larger population centers like this one. Due to the small sample size and low population density of the North East region, there were no significant areas with a high density of cases.⁵ Point data was also reviewed to look for possible patterns in patient demographics, such as patients who were underweight for age; however, no patterns emerged.

This study mapped cases of intestinal perforations due to typhoid fever in the North East region of Ghana, highlighting both the potential and limitations of geospatial analysis in resource-limited settings. Geospatial mapping is an essential tool for targeted public health interventions, particularly in diseases that spread from environmental factors like enteric fevers, cholera, leptospirosis, Hepatitis A, and outbreaks of food contamination. Our study draws parallels with the seminal work of Dr John Snow,¹ but our experience demonstrates how methods that proved revolutionary in 19th century London remain challenging to implement in many resource-limited settings today.

The primary obstacle encountered was the lack of a standardized address system. This deficiency prevented mapping cases to specific neighborhoods or residences, limiting our ability to identify potentially contaminated water sources or implement targeted public health interventions. Although Ghana introduced a national, digital, property-addressing system in 2017, generating unique GPS-based codes, poor uptake in the study region persists because of infrastructural challenges like inconsistent electricity and internet connectivity.⁸ Additionally, the initial promotion of the system focused on its benefits for business and commerce, potentially overlooking its crucial role in health care and public health.⁹ This underutilization of available technology highlights the importance of implementation science in launching large-scale social projects. Policy makers should consider a more holistic approach when implementing such systems, emphasizing benefits across multiple sectors, including health care. Had GPS-based address codes been consistently collected for patients in our study, the precision and impact of our geospatial analysis would have been significantly enhanced.

The small sample size and low population density in the study region further complicated our analysis. While more cases occurred in larger towns, this likely reflects population distribution rather than true case clustering, underscoring the need for larger-scale studies or alternative analytical approaches in sparsely populated areas. Our strict inclusion criteria of only including typhoid fever with intestinal perforations limited our sample size and may have limited our ability to detect meaningful patterns. Despite the lack of case clustering, our results demonstrate that typhoid fever is occurring widely throughout the region and remains a significant health concern, requiring ongoing public health interventions.

Several practical challenges emerged during implementation, including duplicate village names, reliance on unpaved or unofficial roads, preference for landmark-based navigation, and scarcity of published maps. Previous groups have suggested novel approaches to improve small area map analysis at a city and district level; however, these approaches require significant investment and become less feasible when the study area involves multiple cities or districts.^10–13 One potential approach to overcoming this obstacle would be to physically travel to the patients’ homes to obtain GPS coordinates and identify their local water sources. Alternatively, patients with smart phone devices could provide GPS coordinate information from their home during follow-up phone calls.

Our study has several limitations. The reliance on clinical diagnosis rather than laboratory confirmation of typhoid fever may have led to some misclassification. Since the hospital site is a major referral center for the North East area, it creates a geographic sampling bias with patients generally coming from the surrounding region.

Future research should focus on developing and implementing geospatial mapping techniques tailored to the unique challenges of resource-limited settings. This could include innovative approaches to address coding, utilization of mobile technology for data collection, and integration of traditional local knowledge with modern mapping techniques. Additionally, larger scale studies covering wider geographic areas and longer time periods could provide more robust data for hot spot analysis and trend identification.

In conclusion, our study underscores the critical importance of geospatial analysis in understanding and combating infectious diseases such as typhoid fever. Although we successfully mapped most cases to the town level, the lack of standardized addressing systems and basic infrastructure hindered more precise analysis. These findings highlight the urgent need for improved mapping technologies in resource-limited regions, combining technological solutions with local knowledge. This will ensure equitable and effective public health interventions worldwide by contributing to more targeted and efficient health care delivery. Moving forward, increased adoption of digital addressing systems and enhanced infrastructure will be crucial for implementing targeted public health interventions, not only for typhoid fever but for other infectious diseases in similar resource-limited settings worldwide.

GPS

global positioning system

The authors extend special thanks to Drs Poanna Bennam and Nivedha Sukumar from the University of Texas Health Science Center at Houston and to Dr Abbey Myer from Waco Family Medicine for their assistance with data collection.