Improving the leptospirosis disease burden assessment by including ambulatory patients from outpatient departments: a cross-sectional study

Introduction

Assessing the true burden of disease is required for proper health planning and resource allocation, including the control of transmissible diseases such as leptospirosis. Sri Lankan communicable disease burden estimates are usually done using routinely reported data in the surveillance system¹. Lack of actionable diagnostic tests and the diversity of clinical features leading to under-notification of leptospirosis are the major reasons for poor estimation of this disease, a leading cause of acute febrile illness in Sri Lanka^2,3. A recently published systematic review has suggested a correction factor for hospitalized leptospirosis cases to estimate the burden of this disease more accurately. This study estimated the incidence of leptospirosis in Sri Lanka as 52.1 per 100,000 population³. However, these estimations and corrections are made for hospitalized patients without considering outpatient departments (OPDs). It is estimated that approximately 5–15% of outpatients with undifferentiated febrile cases could be due to leptospirosis^4,5, and undifferentiated febrile patients usually present to OPDs. Finally, these estimates have not been applied to assessing disability-adjusted life years, which is always a challenge for acute febrile illnesses. Therefore, prospective studies in the outpatient setting are essential for estimating the burden of disease due to leptospirosis, which in turn is needed to justify investment in diagnostics and vaccine development.

Few studies have assessed leptospirosis in non-hospitalized patients with acute febrile illness. Biggs et al. highlighted the underestimation of leptospirosis due to the non-inclusion of ambulatory patients for disease estimates in Tanzania⁶. A study conducted in Vanuatu showed the importance of screening for leptospirosis among acute febrile illness patients presenting to OPDs during outbreaks, highlighting the need for improved awareness and diagnostic capacity, which are interrelated⁷. In the Vanuatu study, 12 of 161 (7.4%) suspected patients were confirmed as having leptospirosis. However, only 2 of 12 confirmed patients had criteria fulfilling the surveillance case definition, showing the inadequacy of the case definitions used⁷. Another study conducted in Guadeloupe, Martinique (French territories in the Caribbean) suggested that the actual burden of leptospirosis could be 3 to 4 times higher than reported cases⁸. A study conducted in Mozambique also provided supportive evidence for the importance of outpatient leptospirosis by estimating that as much as 10% of febrile patients attending ambulatory care could be attributed to leptospirosis⁹. The purpose of the present study was to determine the prevalence of leptospirosis in an OPD setting in a regional public hospital in Sri Lanka to provide further estimation of disease burden estimations.

Methods

Results

A total of 2,960 febrile patients were screened in the fever section of the OPD during the study period. Of these, 33 (1.1%) were clinically suspected leptospirosis patients and all were recruited for the present study (Figure 1). These included 23 (69.7%) men and 10 (30.3%) women. The mean age was 46.5 years (SD 17.1). During the same period, RPGH made 82 notifications of possible cases of leptospirosis from hospitalized patients. The missing OPD patients from the notification accounted for 28.6% (95% CI 19.4-40.4) (Table 1).

Figure 1. Flow chart of patient selection and diagnosis.

Of 33 possible cases, 8 (24.2%) were laboratory-confirmed as leptospirosis. One patient was categorized as “probable” with a single MAT titre of 1/200¹². Of the 33 cases selected, 12 (36.4%) had received antibiotic treatment from a primary care centre before coming to the RPGH OPD. During the same period, we interviewed 29 hospitalized patients who were treated presumptively for leptospirosis. Of these, 19 (66.5%) reported that they were given treatment for fever from a primary care provider prior to hospital admission. However, none of these 19 visited the OPD of RPGH, confirming that the cases presented to OPD are really “missing” from the system.

Discussion

In this preliminary study to evaluate the missing leptospirosis patient load in the surveillance system, we made three important observations: (1) almost one-third of the patients presenting to the OPD of RPGH were missing from the notification system; (2) most of the patients (although we could say none, there might be admissions after the study period) presenting to the OPD were not hospitalized; (3) most of the hospitalized patients sought healthcare from primary care centres rather than from a tertiary care centre. The OPD data clearly shows that 28.6% (95% CI 19.4-40.4) of leptospirosis patients presenting to this tertiary centre were not included in the system. Nevertheless, statistical assumptions cannot be made for the primary care institution without proper studies conducted in local hospitals and private healthcare institutions. This study mainly focused on the cases presenting in an endemic setting and during an outbreak period. The missing numbers can neither be generalized to all areas of Sri Lanka nor for all the months of the year in the same area. Establishing a well-functioning disease surveillance system in OPDs and primary care institutions is essential for proper disease burden estimates, not only for leptospirosis, but also for other notifiable diseases. Various small-scale studies have been conducted to identify feasible methods for disease surveillance, such as incorporating smartphone technology, which is being carried by hand by the treating physician¹³. These feasibility studies need to be upscaled to identify the barriers and feasible methods to implement the system. Well-planned studies covering outpatient, inpatient, and private sectors should be initiated to estimate the actual burden of diseases.

Data availability