Effect of Supervision or Short Messages (SMS) on the Detection Rate of Diseases Under Surveillance in the Littoral Region of Cameroon: A Randomized Controlled Health Facility-Based Trial

Abstract

Introduction

The Integrated Disease Surveillance and Response (IDSR) strategy was adopted in Cameroon in the early 2000s. However, some limitations in terms of the detection rates of diseases under surveillance are still observed. The present study was conducted to assess the effect of supervision or Short Message Service (SMS) reminders on the detection rate of cholera and measles in health facilities in Douala.

Methods

This was a randomised controlled health facility-based trial that involved assigning health facilities to either monthly supervision of personnel in charge of surveillance, SMS reminders, or a control group. The effect of the interventions was assessed by comparing the detection rates of cholera and measles by personnel with respect to the number of expected suspected cases in consultation records.

Results

A total of 271 health facilities were assessed for eligibility and 159 were included in the trial (53 per study group). The health facilities were assigned to either supervision, SMS or control group using stratified randomisation by health district. The detection rates of suspected cholera and measles cases were 11.05% and 35.19% in the control group, respectively. Health facilities in the supervision group demonstrated significantly higher detection rates of cholera cases [IRR (Incidence Rate Ratio) =1.87 (1.09–3.19); p = 0.022] and measles cases [IRR=1.78 (1.01–3.15); p=0.046], while there was no detected difference in the SMS group compared to the control group for both cholera and measles. The combined detection rate of both cholera and measles was significantly higher in the supervision group than in the control group; however, exposure to SMS did not produce a significant effect. After adjustment, results were consistent, except for detection of cholera cases, where the effect of supervision was attenuated and no longer statistically significant (adjusted IRR=1.39 (0.80–2.42); p=0.243).

Conclusion

Supervision significantly improved the detection of suspected cholera and measles cases, whereas SMS reminders showed no significant effect. Strengthening routine supportive supervision should be prioritised to improve IDSR performance.

Background

Epidemics and pandemics represent a large proportion of morbidity and mortality worldwide.^1,2 Epidemiological surveillance serves as a system for outbreak preparedness and informed decision-making for the control and response to public health threats.^3,4 In sub-Saharan Africa, frequent outbreaks of diseases under routine surveillance, such as cholera, measles, meningitis, and more recently emerging diseases, such as COVID-19, have highlighted persistent gaps in the timely detection and notification of suspected cases of diseases under surveillance.^5–7

The 2005 revision of the International Health Regulations (IHR) centers on strengthening national core capacities and recommends the development of public health surveillance and response systems in low- and middle-income countries to reduce the burden of infectious diseases and prepare for outbreaks.⁸ In the AFRO WHO region, most countries have adopted the Integrated Disease Surveillance and Response (IDSR) strategy, but implementation remains uneven.^9,10 Challenges such as insufficiently trained human resources, weak supervision systems, inconsistent data transmission, lack of financial resources, and delays in laboratory confirmation continue to hinder the performance of the implemented strategy.^9,11,12

In Cameroon, the IDSR strategy was officially adopted in the early 2000s, following recommended guidelines.^13,14 Despite these improvements, the country continues to experience outbreaks of preventable diseases, reflecting weaknesses in the surveillance system.¹³ The most common recent outbreaks included cholera, measles and COVID-19.^15–18 Health facilities in this context are often the first point of contact for patients with the health system and the point of identification of epidemic-prone diseases with respect to local surveillance guidelines.^14,19 Under detection at the health facility level remains a major limit in the surveillance system, coupled with the fact that not all cases do not necessarily reach health facilities due to local health-seeking behaviours.²⁰ Some documented reasons for underperformance of health facilities include poor preparedness of health facilities, limited financial resources, limited laboratory capacity, and work overload.^21–23 These reported low detection rates not only delay the outbreak response but also lead to underestimation of disease burden. The Cameroon epidemiological situation is characterized by recurrent epidemics, with the most recent major outbreaks being cholera and measles.^24,25 The most recent 2018–2023 cholera outbreak in Cameroon resulted in a very important number of reported cases with a case fatality rate of 2.7%.²⁶

Various strategies have been proposed to improve case detection. These include refresher training, supportive supervision, laboratory data exploitation and use of mobile health (mHealth) tools.^27–31 These interventions are particularly appealing in LMICs because of their relatively low costs and ease of deployment. Supportive supervision has been associated with improved health worker performance and surveillance data quality in LMICs and SMS-based reminders have shown promising effect in improving health worker adherence to protocols in similar contexts.^32–34 However, most existing evidence relies on observational or pre-post designs without control groups. Furthermore, no randomized trial has directly compared these two approaches for disease surveillance strengthening in Cameroon at health facility level. This study aims to evaluate the effect of enhanced supervision and Short Message Service (SMS) reminders on the detection rate of cholera and measles in health facilities in the littoral region.

Methods

Study Design

This was a randomised controlled health facility-based trial that involved assigning health facilities to either (i) monthly supervision of personnel in charge of surveillance or (ii) bi-weekly SMS sent to the health personnel in charge of surveillance, reminding them to conduct a search of suspected cases of cholera and measles based on case definitions or to a control group where no study intervention was administered.

Setting

The study was conducted in four health districts in the littoral region, precisely located in Douala. Douala is the economic capital and one of the most densely populated urban areas in Central Africa. This area faces particular outbreak risks due to rapid urbanisation, high population mobility, and low levels of sanitation. The trial was conducted in health facilities of the health districts of Deido, Bonassama, New Bell, and Cité Palmiers.

Participating Health Facilities

Health facilities included in the study were health facilities involved in epidemiological surveillance, had personnel in charge of implementing epidemiological surveillance activities, had consented to participate in the trial during a preliminary health facility visit, and owned a functional mobile telephone.

Intervention

The interventions targeted the surveillance focal points of health facilities that agreed to participate in the study and were implemented over a 06-month period from October 2024 to March 2025. Two interventions were administered in the present trial.

Short Messages (SMS)

Epidemiological surveillance focal points that consented to participate in the trial provided their mobile phone numbers, which were checked by sending a verification SMS to each owner. Pre-established standardised messages translated in English and French were sent to each epidemiological surveillance focal point bi-weekly for six months. The same message was sent during the intervention period. The content of the messages included a reminder to conduct suspected case searches in health facilities’ consultation records, collect the samples, and report identified suspected cases. The content of the message was: Please check the consultation registers for suspected cases of cholera and measles. For any suspected cases, please notify and collect samples to be sent to the laboratory for confirmation.

Supervision

Monthly, independent personnel trained in supervision visited each health facility assigned to the supervision group to supervise the focal point on suspected cases’ detection and reporting of identified cases. Weaknesses on case definitions were corrected by supervisors using standardised guidelines. Key points of supervision included reminders and revision of the definition of suspected cases, consideration of the frequency of searching for suspected cases in consultation records, and the difficulties encountered.

Control

No intervention was administered to health facilities assigned to the control group.

Sample Size

The sample size was estimated at 150 health facilities (50 in each study group) assuming there will be a 5% increase in the suspected case detection rate between supervision or SMS groups compared to the control, 80% statistical power, and a 0.05 level of significance with a 2-tailed test.

Randomisation

The sampling frame of health facilities was constructed by independent field surveyors responsible for facility recruitment. Participating health facilities were allocated to three groups (supervision, SMS, and control) using stratified randomisation by health district. Stratification was performed by health district to account for differences in the number of health facilities per district. Within each district, health facilities were assigned to the study arms using a centralised computer-generated randomisation sequence prepared prior to the start of the intervention. The sampling frame of eligible health facilities was established by surveyors responsible for facility recruitment and geolocation. The randomization sequence was generated independently by a team member not involved in facility recruitment, using an appropriate software (sealed envelope).

Outcome Estimate

To assess the effect of the interventions on the detection rate of the targeted diseases, the number of suspected cases per disease was counted by independent trained surveyors in health facility consultation records for the period covering the intervention administration period (October 2024 to March 2025). The suspected cases were identified based on standardized case definitions used nationally. During the same period, the number of suspected cases identified per focal point was recorded. The primary outcome was the detection rate of suspected cholera and measles cases that was collected and compared between the study and control groups. This was determined using the number of cases collected in the health facility records as denominators and the number of suspected cases detected by health personnel over the same period.

Data Management and Analysis

All missing values were crosschecked and cleaned before data analysis. Descriptive statistics were used to summarise the characteristics of the study population. Categorical variables are presented as frequencies and percentages. The incidence rate of suspected cases, which corresponds to the detection rate, was estimated using a Poisson regression model with an offset for the expected number of cases, and predicted incidence rates by study arm were derived using marginal standardization. The Poisson regression model with the log of expected cases as an offset was used to estimate the incidence rate ratios (IRRs) and their 95% confidence intervals (CIs). Adjusted analysis to estimate adjusted incidence rate ratios (aIRR) were conducted with respect to health districts, type of health facilities and category of health personnel. Data were analysed using Stata version 18, and statistical significance was set at p < 0.05.

Ethical Considerations

All participants were informed of the objectives and procedures of the survey and only those who consented to participate were included. No confidential or personal data were collected. All methods related to this study were performed in accordance with the Declaration of Helsinki. The research protocol of the study was approved by the Littoral Regional Ethics Committee for Human Health Research (N°2023/0017/CE/CRERSH-LITTORAL).

Results

Recruitment and Participant Flow

Before study initiation, 271 health facilities in the four targeted health districts were visited to assess eligibility and seek consent to participate in the study. A total of 217 (83.0%) health facilities were eligible and 159 (70.7%) consented to participate. Figure 1 presents the enrolment and inclusion of health facilities during the study process.

Figure 1.

Flow diagram of inclusion of the study.

Characteristics of Participating Health Facilities and Health Personnel

Table 1 presents the characteristics of the health facilities and health professionals by study groups. A total of 127 (92.0%) of the participating health facilities where health centers, 68.8% of participants were nurses. The health facility category, category of health personnel receiving the intervention, type of health facility, and type of health district were distributed evenly across the three study groups.

Table 1.

Characteristics of Health Facilities and Health Professionals by Study Groups

Characteristics	Supervision Group		SMS Group		Control Group		P-value
	n	%	n	%	n	%
With respect to health facility category
Health centres	44	34.6	41	32.3	42	33.1	0.364
Sub-divisional medical centers	0	0.0	4	57.1	3	42.9
District level hospital	2	50.0	1	25.0	1	25.0
With respect to category of health personnel receiving the intervention
Biologist	2	22.2	4	44.4	3	33.3	0.475
Epidemiologist	0	0.0	1	33.3	2	66.7
Nurse	32	33.7	32	33.7	31	32.6
Physician	12	41.4	9	31.0	8	27.6
Pharmacist	0	0.0	0	0.0	2	100.0
Type of health facility
Private confessional	5	20.8	11	45.8	8	33.3	0.410
Private	39	36.1	32	29.6	37	34.3
Public	2	33.3	3	50.0	1	16.7
With respect to the type of health district
Urban	31	34.1	29	31.9	31	34.1	0.879
Semi-urban	15	31.9	17	36.2	17	36.2

Detection Rate of Suspected Cases of Targeted Diseases Under Surveillance

The detection rates of suspected cholera cases in the supervision, control, and SMS groups were 20.67%, 11.05%, and 8.94%, respectively. Table 2 presents the detection rates of suspected cases of cholera, measles, and cholera/measles per study group. Table 3 presents the absolute number of suspected cholera and measles cases detected and expected by study arm.

Table 2.

Detection Rate of Suspected Cases of Cholera and Measles per Study Groups

Study group	Detection Rate (%)	95% CI
Cholera
Control	11.05	[6.33–15.78]
Supervision	20.67	[14.01–27.33]
SMS	8.94	[5.11–12.76]
Measles
Control	35.19	[19.36–51.00]
Supervision	62.75	[41.01–84.48]
SMS	42.11	[25.26–58.95]
Cholera and Measles
Control	16.39	[11.31–21.17]
Supervision	30.00	[22.92–37.08]
SMS	15.41	[10.91–19.91]

Table 3.

Absolute Number of Suspected Cholera and Measles Cases Detected and Expected by Study Arm

Study Group	Cases Detected	Expected Cases
Cholera
Control	21	190
Supervision	37	170
SMS	21	235
Measles
Control	23	54
Supervision	35	51
SMS	36	57
Cholera and Measles
Control	44	244
Supervision	72	230
SMS	57	292

Contribution of Supervision or SMS on Cholera and Measles Detection Rate

Table 4 presents the incidence rate ratios comparing the intervention groups with respect to the control group. The detection rate of cholera suspected cases in the supervision group was superior to that of the control group [IRR = 1.87 (1.09–3.19); p = 0.022)], while there was no detected difference between the detection rate of cholera in the SMS group and that of the control group [IRR = 0.81 (0.44–1.48); p=0.491]. The detection rate of measles was higher in the supervision group compared to control group [IRR = 1.78 (1.01–3.15); p = 0.046], whereas there was not detected difference in the SMS group [IRR = 1.19 (0.66–2.18); p = 0.559]. The combined detection rate of cholera and measles was significantly higher in the supervision group than in the control group, but exposure to SMS did not significantly improve the detection rate. After adjustment, results were consistent, except for suspected cholera cases, where the effect of supervision was attenuated and no longer statistically significant (adjusted IRR=1.39 (0.80–2.42); p=0.243).

Table 4.

Incidence Ratio Between Intervention Groups and the Control Group

	Crude			Adjusted
	IRR	95% CI	P-value	IRR	95% CI	P-value
Cholera
Supervision	1.87	[1.09–3.19]	0.022	1.39	[0.80–2.42]	0.243
SMS	0.81	[0.44–1.48]	0.491	0.84	[0.43–1.63]	0.608
Control	1	–	–	1	–	–
Measles
Supervision	1.78	[1.01–3.15]	0.046	2.21	[1.17–4.16]	0.014
SMS	1.19	[0.66–2.18]	0.559	1.31	[0.67–2.56]	0.429
Control	1	–	–	1	–	–
Cholera and measles
Supervision	1.83	[1.24–2.70]	0.002	1.51	[1.01–2.25]	0.047
SMS	0.94	[0.61–1.44]	0.776	0.81	[0.52–1.28]	0.366
Control	1	–	–	1

Discussion

This study was conducted to assess the effect of supervision or short messages (SMS) on the detection rate of suspected cases of diseases under surveillance with cholera and measles as case studies. The detection rates of suspected cholera and measles were 11.05% and 35.19%, respectively, in the control group. The detection rate of cholera suspected cases in the supervision group was superior to that of the control group [IRR = 1.87 (1.09–3.19); p = 0.022)], while there was no detected difference between the detection rate of cholera in the SMS group and that of the control group [IRR = 0.81 (0.44–1.48); p=0.491]. However, after adjustment, the detection rate of cholera suspected cases in the supervision group was not statistically superior to that of the control group [IRR=1.39 (0.80–2.42); p=0.243).] The detection rate of measles was higher in the supervision group compared to control group [IRR = 1.78 (1.01–3.15); p = 0.046], whereas there was not detected difference in the SMS group [IRR = 1.19 (0.66–2.18); p = 0.559]. The combined detection rate of cholera and measles was significantly higher in the supervision group than in the control group, but exposure to SMS did not significantly improve the detection rate.

The early detection of diseases under epidemiological surveillance is one of the key objectives of surveillance systems. Case detection by the system in place is the priority and starting point of the epidemiological surveillance cascade.^35–38 The detection rates of suspected cholera and measles in the control arm of the present study that represents routine practice were 11.05% and 35.19%, respectively. This detection rate is low compared to the 80% benchmark for evaluating the performance of surveillance system indicators.³⁹ These detection rates suggest that routine case surveillance may miss a substantial proportion of the suspected cases. These findings were consistent with those of previous studies conducted in different settings. Previous studies have documented underreporting of cholera and underestimation of the actual disease burden.³⁵ Similarly, underreporting of measles has been documented, where a substantial proportion of suspected cases have never been reported or investigated.⁴⁰ The low detection rates could call into question the effective practice of routine surveillance alongside other activities in health facilities, case definition applications, reporting practices, or health system responsiveness. Some contributing factors to these limits in terms of surveillance include; with respect to settings, limited resources, workload, lack of or limited laboratory capacities, health system instability, and poor data-sharing systems.^41–43 These results suggest the need to identify measures to improve disease surveillance.

Strengthening disease surveillance in low- and middle-income countries remains a public health priority with respect to the documented weaknesses. Low-cost interventions such as mobile phone reminders and supervision have been increasingly explored to address these gaps. The present study assessed the effect of including supervision or SMS in surveillance on the detection rate of cholera and measles, diseases to which Cameroon is constantly exposed. The study demonstrated that monthly supervision significantly improved the detection of cholera and measles cases, whereas SMS reminders did not show a significant effect. However, after adjustment, the effect of supervision on suspected cholera cases detection alone was attenuated and no longer reached statistical significance (adjusted IRR=1.39 [0.80–2.42]). The beneficial effect of supervision is consistent with evidence from other surveillance and health system strengthening strategies in Cameroon and other settings, where supervision has been shown to improve data quality, reporting, health workers’ performance, clinical practices, and reporting rates of adverse events following immunisation.^32,34,44 Similar to previous investigations, the positive performance of supervision can be explained by the interactive nature of this intervention. Supervision visits included direct feedback, propositions of solutions to identified problems, and refreshments of surveillance case definitions. This interactive nature may explain the boost in health personnel motivation and adherence to cases’ definitions. However, it is difficult to determine which component contributed most to the observed effect. While supervision is effective, it may require substantial financial and human resources, which may limit its feasibility and sustainability at scale in resource-limited settings. Future studies should formally assess the cost-effectiveness of this approach.

In contrast to the effect of supervision, SMS reminders did not have a significant effect on the detection rates of suspected cholera and measles cases. Previous studies that assessed the effect of including short messages to improve system performance had mixed results. Most studies have shown the positive effect of SMS reminders on the improvement of reporting, timeliness, data completeness, and transversal tasks.^33,45–47 Some studies found a smaller effect on indicators, such as detection rate, but the effects appeared to be more significant when combined with other approaches, including training, incentives, and supervision.^34,48–50 The lower effect of SMS reminders may be explained by the fact that receiving a text message does not in any way trigger a direct obligation at the moment or increase motivation, and these reminders could be deprioritised in the face of other competing tasks to be carried out by health personnel. This could also be due to message fatigue and lack of interaction with the health personnel. This one-way communication alone may not be sufficient to change the established practices. The detection of diseases under surveillance is also heavily associated with health personnel preparedness regarding the knowledge of surveillance procedures, recognition of cases, and case definitions. Unlike the implementation of supervision, which included training on surveillance guidelines and case definitions, the use of SMS only reminded them to perform case surveillance but did not improve their ability to recognise cases. This may also explain the difference and the need to associate SMS reminders with other improvement factors.

This study had some limitations. The study was conducted in only four health districts and did not cover all the categories of health facilities in Cameroon; thus, it may not be generalisable to all regions. There was no guarantee that all health personnel relied on the same case definitions for detection, which could have influenced the number of cases detected. Data on some variables that could have been included in the regression model, such as past training of the health personnel in charge, seniority level and laboratory capacity, were not collected. In addition, the study was conducted over a period of six months, this relatively short period may not be sufficient to assess whether the observed improvements could be sustained in the long term. The relatively low number of suspected cases detected may have limited statistical precision resulting in the observed relatively wide confidence intervals. Larger studies would be needed to confirm the magnitude of the observed effects. The intervention was implemented in a city with better health infrastructure and relatively stable network coverage; thus, SMS design may not be feasible in remote settings with no network coverage.

Conclusion

Supervision was associated with a higher detection rate of suspected cases of cholera and measles, whereas SMS did not have a significant effect. Investments in supervision should be prioritised when resources are available, and SMS can be considered a complementary tool for other interventions rather than a standalone intervention.

Funding Statement

This research received no specific grants or external fundingfrom any funding agency.

Data Sharing Statement

The datasets used and/or analysed during the current study are available from the corresponding author upon request.

Ethical Approval and Consent to Participate

Only participants who provided informed consent were included in this study. The study protocol was approved by the Littoral Ethics Committee for Human Health Research of Cameroon (N^o2023/0017/CE/CRERSH-LITTORAL).

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis, and interpretation, or in all these areas, took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no competing interests in this work.