Introduction

Malaria is a major public health concern that requires quality data management systems for effective decision making towards planning, and design of preventive measures. The World Health Organization (WHO) report on malaria shows that 244 million and 249 million cases of malaria were reported in 2021 and 2022 respectively1. Malaria accounted for 608 000 deaths in 2022 compared to 610 000 deaths in 20211. The incidence of malaria has increased over the years with significant burden in sub-Sharan African countries which accounts for approximately 94% of all malaria cases globally1. Malaria in children under 5 years of age is 78% of all malaria mortalities in the African region1,2.

Malaria is endemic in Ghana and accounts for approximately 40% of outpatient department cases3,4. Malaria in Ghana exhibit spatial heterogeneity, with the burden varying significantly across the different regions. Factors such as access to quality health care, seasonal variations, and the implementation of malaria preventive and control measures including the use of long-lasting insecticide nets (LLINs) influence the transmission and prevalence of malaria5,6,7,8. The prevalence of malaria in Ghana saw a significant reduction from 27.5% in 2011 to 8.6% in 20229. Similarly, confirmed malaria cases per 1000 population declined from 192 in 2019 to 159 per 1000 in 20209. Despite the significant decline in malaria prevalence over the years, it still remains unacceptably high and a major national health burden, with 20.9% prevalence in the Shai Osuduko Municipality in the Greater Accra Region10 and 73% of asymptomatic Plasmodiuminfection in the Asante Akim North district in the Ashanti region11. Pregnant women and children are the most vulnerable to malaria infections with some Ghanian studies reporting the prevalence of malaria in pregnancy to range between 8.9–14.1%12.

A national representative demographic health survey reported the prevalence of malaria in children under 5 years to be 22.1%13. To effectively eliminate the burden of malaria and achieve targeted distribution of preventative measures, the importance of routine quality data cannot be downplayed14,15. High quality and reliable data are important for malaria epidemiological studies, monitoring and evaluation of interventions and treatment programs16,17. In Ghana, the District Health Management Information System (DHMIS II) has been routinely used to capture and aggregate data across all public health facilities18. The DHMIS II facilitates easy access to health care data for analysis at the national, district and community levels and forecast essential healthcare services to inform planning, performance evaluation and data-driven decision making18. The DHIMS II is used in all Ghana’s four-level healthcare system, comprising the health post and Community-based Health Planning and Services (CHPS) zones, health centers and clinics, district, regional and teaching hospitals with varying mandates in health service provision19.

The CHPS zones provide primary health care services including community level health promotion, prevention and primary clinical care at the community and most deprived settings in Ghana20. The CHPS zones serves as the initial point of entry to the health care system and is used to diagnose, treat and manage malaria cases in Ghana20. The CHPS zones are managed by trained community health officers (CHOs) with some level of capacity to manage malaria cases. However, previous studies have shown that CHPS still face substantial challenges including administrative, and minimal training to provide patient care for febrile illnesses such as malaria21. Tracking malaria guidelines remain a challenge in most CHPS zones21. In spite of the challenges faced by CHPS and the role in malaria prevention, management and elimination, the quality of the routine malaria data captured remains a grey area. In an effort to eliminate malaria in Ghana, the Ministry of Health and the Ghana Health Service, with support from partners, launched the National Malaria Elimination Strategic Plan (NMESP) 2024–202822. Under the theme “Zero malaria starts now: Launching Ghana’s path to elimination”. The plan emphasizes the importance of ensuring data quality as a key component of the elimination strategy. The study therefore aimed at assessing the quality of routine malaria data captured on the DHIMS II at CHPS zones in the Hohoe Municipality, where confirmed malaria cases ranged between 20%23to 59.6%24. Findings of the study provides a comprehensive and appropriate information about the quality of malaria data captured on DHIMS II at primary care centers for use by health policymakers. The findings of the study are important for improving the quality of malaria data and guidance to the implementation of measures that addresses factors affecting data quality in DHIMS II at the CHPS zones.

Methods

Study design and setting

The study employed a retrospective cross-sectional study design to conduct routine malaria data quality audit. The study was conducted in CHPS compounds in the Hohoe Municipality, Volta Region, Ghana. The Hohoe Municipality is one of the 18 municipalities located in the central part of the Volta Region in Ghana. The Hohoe Municipal has a population of 114,472, according to the Ghana 2021 population and housing census25. Among the population, females constitute 52.0%, while males make up 48.0% (Hohoe District Health Directorate). Majority of the population (55%) are engaged in engaged in farming activities, with others in trading (25%), livestock rearing (15%), and 5% in other activities (Hohoe Municipal Assembly Profile, 2023). The Hohoe Municipal has 4 sub-health directorates comprising Alavanyo, Agumatsa, Gbi-Urban, and Hohoe-sub. In all, there are 15 health institutions in the Municipality including 4 CHPS compounds.

Data source and extraction

A 5-year (1st January, 2018 to 31st December, 2022) data was extracted from the health facilities registers such as the consulting room register, monthly malaria report registers, and DHIMS II. An excel proforma was designed to extract the data from the health registers of the CHPS compounds. Only records of patients who had malaria within study period data was extracted, and cleaned. Patients with key missing data were dropped and excluded from the analysis.

Study variables

The study extracted variables such as months and year report on malaria, availability of monthly morbidity reports, availability of anti-malarias returns, expected and completed cells on monthly morbidity reports, DHIMS monthly value on malaria based on the indicators suspected malaria cases, suspected malaria cases tested and suspected malaria cases tested positive.

Statistical analysis

Descriptive statistics was performed and a comparison of the information obtained from the two sources to detect any disparities. Frequency tables and bar charts were used to describe the quality of malaria indicators. Calculations were made to determine the percentage of missing data and the percentage of data errors. The assessment of the OPD routine malaria data’s quality involved evaluating the availability, completeness (percentage of missing data), and accuracy (verification factor) of the predefined malaria indicators.

Availability

The accessibility of malaria indicators was assessed by their availability and readiness when needed. This was done by assessing OPD registers and monthly morbidity returns forms. The assessment of malaria data’s availability at all the CHPS centers was based on the accessibility of the required registers and reports. As per the standard monthly malaria summary report outlined by the Ghana Health Service (GHS), achieving 100% report availability entails having all reports (12 in total per year) accessible and ready within the reporting year. To meet the report availability criteria, facilities must surpass the target of ≥ 90%26.

Availability was calculated as:

$$\:\frac{Number\:of\:reports\:received\:from\:all\:participating\:facilities}{Number\:of\:reports\:expected\:from\:all\:the facilities\:}\:\:\times100$$

Completeness

Completeness means that an information system from which the results are derived is appropriately inclusive: it represents the complete list of eligible persons or units and not just a fraction of the list.

Completeness was calculated as follows:

$$\:\frac{Number\:of\:reports\:that\:are\:complete\:from\:all\:the\:facilities}{Number\:of\:reports\:expected\:to\:be\:complete\:from\:all\:the\:facilities}\times100$$

Accuracy

A comprehensive validation process was carried out at the CHPS compounds by selecting critical malaria indicators and cross-referencing the data elements present in the primary records for all CHPS compounds spanning from January 2018 to December 2022. Subsequent validation checks were performed by consolidating the summary reports submitted to DHIMS II. The Verification Factor (VF) serves as the pivotal measure for evaluating the reported data’s quality by comparing it to the source data, which includes the register or other DHIMS II records at the service delivery point. The VF’s interpretation and the degree of data accuracy can be found below27 (Table 1).

$$\:\frac{\:Recounted\:number\:of\:events\:from\:source\:documents}{Reported\:number\:of\:events\:submitted\:to\:DHIMS}\times100$$
Table 1. Accuracy key. Source2728,
Full size table

Ethical issues

The study protocols were approved the University of Health and Allied Sciences Research Ethics Committee (UHAS-REC A.10 [011]22–23). Due to the retrospective nature of the study, the University of Health and Allied Sciences Research Ethics Committee waived the need of obtaining informed consent. However, administrative approvals were obtained from the selected health facilities. Researchers complied with the guidelines of using secondary data in the health facilities. Such regulations included adherence to data security and protection, transparency, privacy, and confidentiality. The data retrieved was anonymized and maintained with no identifiable information.

Results

Availability of source data

Figure 1 shows the availability of source data (Monthly OPD Morbidity Report) from all the CHPS facilities between 2018 and 2022. Three (75%) out of the 4 facilities consistently had source data available for all reviewed years and exceeded the target of (≥ 90%). On the other hand, data was only available for 2 (50%) health facilities in 2021 and 2022, but not in 2018, 2019, and 2020. However, the source data for 2 (50%) health facilities met the target of (≥ 90%) for availability in 2020 and 2021 respectively.

Fig. 1
figure 1

Availability of monthly OPD morbidity report in consulting room registers requested by Community-based Health Planning and Services (CHPS).

Full size image

Figure 2 shows the availability of source data (Monthly Data Returns on Anti-Malarial) in all CHPS facilities. The source data was only available in 2021 and 2022 for all the CHPS facilities in the municipality. Furthermore, the available source data for these years thus 2021 and 2022 exceeded the target of (≥ 90%) for availability, indicating that the facilities successfully met the required data reporting threshold during those years.

Fig. 2
figure 2

Availability of monthly data returns on anti-malarial in consulting room registers requested by Community-based Health Planning and Services (CHPS).

Full size image

Completeness of data

Figure 3 presents the completeness of source data (Monthly OPD Malaria Cases Suspected) from all the CHPS facilities from 2018 to 2022. The data shows that 2 (50%) of the health facilities had source data completed for all the years (2018 to 2019) but they did not exceed the target of (≥ 90%). On the other hand, data was complete for 2 (50%) health facilities in 2021 and 2022 respectively. The source data for one facility in 2020 and 2021 did meet the target of (≥ 90%) making the availability of completed data for those years.

Fig. 3
figure 3

Completeness of monthly OPD malaria cases suspected in consulting room registers requested by Community-based Health Planning and Services (CHPS).

Full size image

Figure 4 shows the completeness of source data (Monthly OPD Malaria Cases Suspected Tested) from all the CHPS facilities in the Hohoe municipality between 2018 and 2022. The data shows that 50% had source data completed for all the reviewed years, but they did not exceed the target of (≥ 90%). Additionally, data was completed for 50% in 2021 and 2022. The figure shows that the source data for 25% of CHPS facilities in 2021 and 2022 met the target of (≥ 90%), indicating the availability of completed data for those years.

Fig. 4
figure 4

Completeness of monthly OPD malaria cases suspected tested in consulting room registers requested by Community-based Health Planning and Services (CHPS).

Full size image

Figure 5 presents the extent of completeness in source data (Monthly OPD Malaria Cases Suspected Positive) for all the CHPS facilities from 2018 to 2022. The data shows that 50% had complete source data for all the years under review, but they did not surpass the target of (≥ 90%). Additionally, 50% had complete data for the years 2021 and 2022. The source data for 25% of the facilities in 2018 and 2022 met the target of (≥ 90%) which indicates data availability.

Fig. 5
figure 5

Completeness of OPD malaria cases suspected tested positive reports requested by Community-based Health Planning and Services (CHPS).

Full size image

Accuracy of source data

Data accuracy of malaria cases suspected from the DHIMS II and consulting room registers

Two (50%) of the health facilities showed accurate reporting with neither over-reporting nor under-reporting and their corresponding data accuracy level was classified as Good (± 5%). On the other hand, the remaining half (of the facilities 2 out of 4 under-reported their suspected malaria cases, and their corresponding data accuracy level was considered Poor (± 11–20%) (Table 2).

Table 2 Data accuracy of malaria cases suspected from the DHIMS II and Consulting Room registers.
Full size table

Data accuracy of malaria cases suspected tested from the DHIMS II and Consulting Room Register

Two (2) out of 4 facilities demonstrated accurate reporting without either over-reporting or under-reporting, and their corresponding data accuracy level was categorized as Good (± 5%). Conversely, the other half of the facilities under-reported their suspected malaria cases tested, and their corresponding data accuracy level was classified as Poor (± 11–20%) (Table 3).

Table 3 Data accuracy of malaria cases suspected tested from the DHIMS II and Consulting Room registers.
Full size table

Data accuracy of suspected malaria cases tested positive from the DHIMS II and Consulting Room Register

Data accuracy of suspected malaria cases tested positive from the DHIMS II and Consulting Room Register showed that half of the facilities 2 (50%) showed accurate reporting without either over-reporting or under-reporting, and their corresponding data accuracy level was categorized as Good (± 5%). On the other hand, the remaining half of the facilities 2 out of 4 under-reported their suspected malaria cases tested positive, and their corresponding data accuracy level was classified as Poor (± 11–20%) (Table 4).

Table 4 Data accuracy of malaria cases suspected tested positive from the DHIMS II and Consulting Room registers.
Full size table

Discussion

The study assessed the availability, completeness and accuracy of routine malaria data captured in DHIMS II and consulting room registers. Consistently, half (50%) of the health facilities consulting registers had reliable data that exceeded the ≥ 90% target for all the 5 years whereas the remaining half (50%) had accessible data in only 2021 and 2022 respectively. The study showed that Monthly Data Returns on Anti-Malarial were accessible in only half of the facilities in 2021 and 2022, meeting the threshold across all the CHPS facilities. Most facilities faced challenges in data availability due to the lack of health data infrastructure and poor health data management systems. The study findings are consistent with that of Okyere-Boadu et al27. which found that 93.3% of the 15 health facilities surpassed the targeted report availability threshold of ≥ 90%. The study equally noted that most of the facilities had all their reports not just accessible but also fully completed27. This could be attributed to the similarities that may exist in the data capturing systems across healthcare facilities. A similar study showed that registers and report forms were the most commonly available source of data and used health management information systems tools in healthcare facilities with high variation between levels of the health system26. The data availability rates of report forms submitted to the district health offices indicate weakness in the transmission or storage of reports at the district level29. The reason for most health facilities reaching the threshold could be attributed to substantial infrastructure and human resource investments for DHIMS II implementation including continuous training, education and support. Moreover, reasons for the missing or inaccurate related data could be largely be attributed to health staff shortages, poor health information infrastructure and high variation between levels of the healthcare system. Though the burden of malaria has seen a significant decline over the years across the globe, however, it still remains a major health challenge in endemic regions and therefore routine health information systems is essential for monitoring progress towards malaria elimination30.

For suspected malaria cases, about half of the health facilities had their source data (registers) filled out for all the reviewed years, though they did not meet the target. The remaining half of the facilities had completed data for 2021 and 2022 respectively. However, less than half of the facility’s data for 2021 and 2022 met the target; an indication that much effort is still needed to achieve data availability for data driven decision making. Likewise, another 75% of the facility’s data did not meet the target for the available years (2019–2022). For monthly OPD-suspected malaria cases, about half of the health facilities had complete source data for all the years under review but they did not meet the target. The complete data is important for the review of historical malaria trends that can support decision-making and malaria prevention strategies. Not achieving the recommended target al.so presents other opportunities for training on data capturing and management. The current study’s finding is inconsistent with the findings of Githinji et al31. which shows that the reporting completeness for confirmed malaria cases, artemether–lumefantrine (AL) treatments, long-lasting insecticidal bed nets (LLINs) distributed via antenatal care (ANC) and child health clinics improved significantly in the Lake endemic zone. The observed difference may be partly attributed to an adoption of a more rigorous process of data collection and validation before transfer from one medium to another32. Likewise, variations in geographic locations and the operations of the health services could contribute to the observed findings.

Accurate measurement of malaria indicators is important to track progress made and develop targeted interventions towards malaria elimination33,34. The current study’s findings showed that half of the facilities (50%) demonstrated very high data accuracy. Prioritizing data accuracy places healthcare facilities in positions to affect healthcare decision makings positively. Previous studies have shown that accuracy of malaria case reporting remains a challenge in most countries across the globe33,34,35. The current study equally found that some of the health facilities showed significant inaccuracies in the years under examination. The Ghana Health Service report shows malaria data quality has been a challenge particularly at the sub-national levels36. A Nigerian study reported similar findings where data quality for district health information system for malaria was found to be suboptimal37. According to Okello et al38., the challenges to routine malaria data generation at major healthcare facilities are not malaria specific and highlights the general weakness in the healthcare system weaknesses. Providing healthcare workers and data managers at the primary care levels with the essential knowledge and skills in malaria routine data quality is important to improve data accuracy39. Collaborating the findings of Sychareun et al40. showed that data on the six indicators gathered and reported in the public health system in the provinces of Xiengkhouang and Houaphanh lacked accuracy at health centers and district health offices. This could be as a result of inaccurate number recording, the monthly report not included all registers, or a compilation of register data before the end of the month, which leaves out some data. Again, the health management information system may lack uniform reporting style, resulting in several international funding agencies to utilize distinct data sources and create their own reports40.

Limitation

The research utilized information obtained from the DHIMS II database, monthly morbidity reports, and the consulting room register. The data collection specifically targeted the assessment of variables’ availability, completeness, and accuracy within the database. However, the analysis did not involve patient-specific details, and interviews from healthcare workers. The absence of information regarding patient history, malaria risk factors, diagnostics, and treatments limits the study for the purpose of malaria interventions, however the study highlights the need for quality routine malaria data at primary healthcare facilities. Similarly, the study was conducted in only one municipality out of 18 administrative districts of the Volta Region. The use of the study’s findings should be contextualized and with caution.

Conclusion

The results of the study shows that although half of the facilities had source registers with complete and adequately available malaria data, there are discrepancies with the information presented in the DHIMS II database. Data availability and completeness metrics in all facilities were consistently below recommended standards. However, the accuracy was moderately commendable. Most of the facilities involved in the study revealed significant issues with data quality, such as incomplete data, inaccurate data, inconsistent data, missing values, and data entry errors. Although some data were deemed reliable for day-to-day decision-making in healthcare, there remains room for enhancement in their quality, particularly when making comparisons across facilities.