Outbreak of leptospirosis in New Caledonia: diagnosis issues and burden of disease

Authors


Corresponding Author C. Goarant, Institut Pasteur de Nouvelle-Calédonie, BP 61, 98845 Nouméa Cedex, New Caledonia. Tel.: +687 27 97 46; Fax: +687 27 33 90; E-mail: cgoarant@pasteur.nc

Summary

A leptospirosis epidemic affected New Caledonia during the first semester of 2008. A total of 135 cases were diagnosed with a relatively low fatality rate of 3.7%. Heavy rainfalls, related to La Niña, favoured this epidemic. The PCR, routinely used, confirmed 54% of the cases, and the microagglutination test 56%. Epidemiological and economical data on this epidemic are presented and discussed.

Abstract

Epidémie de la leptospirose en Nouvelle-Calédonie: enjeux du diagnostic et charge de morbidité

Une épidémie de leptospirose a affecté la Nouvelle-Calédonie au cours du premier semestre de 2008. Un total de 135 cas ont été diagnostiqués avec un taux de létalité relativement faible de 3,7%. Des pluies abondantes liées à‘La Niña’ ont favorisé cette épidémie. La PCR utilisée en routine a confirmé 54% des cas et le test de microagglutination 56% des cas. Les données épidémiologiques et économiques sur cette épidémie sont présentées et discutées ici.

Abstract

Brote de leptospirosis en Nueva Caledonia: cuestiones diagnósticas y carga de enfermedad

Una epidemia de leptospirosis afectó a Nueva Caledonia durante el primer semestre del 2008. Se diagnosticaron 135 casos con una tasa de fatalidad relativamente baja del 3.7%. Las lluvias torrenciales, relacionadas con La Niña, favorecieron esta epidemia. La PCR, utilizada rutinariamente, confirmó un 54% de los casos, y la prueba de microaglutinación un 56%. Se presentan y discuten los datos epidemiológicos y económicos sobre esta epidemia.

Introduction

Leptospirosis is the most widespread zoonosis occurring worldwide, possibly lethal, with epidemics associated with meteorological events, notably under the influence of the El Niño Southern Oscillation (ENSO) (Levett 2001). Its recent emergence or re-emergence (Bharti et al. 2003; Pappas et al. 2008) points to the lack of preparedness in low endemic locations, where epidemics may occur. There notably is a recognized need for a better assessment of its real incidence, medical and sociological impacts and costs in endemic or epidemic contexts (Smythe 1999; Hartskeerl 2005). In New Caledonia, leptospirosis is known to be endemic with epidemic bursts occurring during hot rainy periods (Merien & Perolat 1996; Berlioz-Arthaud et al. 2007a,b). In the first semester 2008, a La Niña situation induced heavy rainfalls and floods and was associated with an epidemic of human leptospirosis, causing 135 cases and 5 deaths.

Study methods

The population studied was all lab-confirmed cases of leptospirosis identified during the first half of 2008. All leptospirosis suspicions were asked a questionnaire about possible exposures, clinical signs and geographical localization. Medical and social data [number of days of hospitalization, including in intensive care units (ICU) and loss of workdays including days of hospitalization] were collected for every patient hospitalized. The standard cost of a hospitalization day in a peripheral health center, a standard unit or ICU was obtained from the health centers and hospital respectively. For patients whose condition did not require hospitalization, a telephone inquiry of clinicians allowed to define a standard analysis and treatment regimen (leptospirosis diagnosis, hematology, biochemistry and a routine electrocardiogram) and an average loss of workdays per patient. The economical results were then extrapolated to 1 year based on the total number of leptospirosis cases diagnosed in 2008. The geographical origin of the patients was collected from either the return forms or by a telephone inquiry of the patients or their doctors.

The diagnosis was made at the Pasteur Institute of New Caledonia, the central reference laboratory for human leptospirosis. Both early molecular diagnosis, using a real-time PCR procedure with sera or urines (Merien et al. 2005) and late serological diagnosis using a panel of Leptospira strains of epidemiological relevance with the microagglutination technique (MAT) (Berlioz-Arthaud et al. 2007a) were used. Leptospirosis cases were categorized as confirmed (a positive PCR or a sero-conversion in paired samples) or probable (having both a clinical presentation of leptospirosis and a single MAT titre ≥ 800) (Berlioz-Arthaud et al. 2007a).

Results

A total of 135 confirmed (101) or probable (34) leptospirosis cases were diagnosed during the first half of 2008. This number of cases, though lower than in large epidemics in 1997 and 1999 (Berlioz-Arthaud et al. 2007b) is much higher than the yearly average (Merien & Perolat 1996; Berlioz-Arthaud et al. 2007b) reflecting an epidemic most probably related to higher rainfalls and floods under the influence of a ‘la Niña’ phase of the ENSO. The distribution of cases and pluviometry data are shown in Figure 1. The mean age of cases was 35.2 years (4.6–84.3) and most were males (M/F = 2.2). Symptoms are summarized in Table 1.

Figure 1.

 Geographic distribution of the incidence of leptospirosis during the first semester 2008: evidence of a rural epidemic. Insert: Rain in Bourail during the period, as a percentage of the monthly average over the past 10 years (data kindly provided by Meteo France) and leptospirosis cases in New Caledonia.

Table 1.   Symptoms reported from lab-confirmed leptospirosis and negative cases
 FeverHeadacheMyalgiaConjunctival suffusionsIcterusRenal syndromeHaemorrhagic syndromeMeningeal syndromeCardiac syndrome
98 cases of leptospirosis91%86%86%34%31%23.5%15.3%9.2%9.2%
410 negative cases82%72.8%61.9%13.4%5.9%4.6%3.8%2.5%1.5%

Early diagnosis using qPCR detection of the aetiological agent in biological fluids allowed the diagnosis of 54% of the 135 cases, serological diagnosis identified 56% of these cases.

The qPCR technique used, together with the fact that only two genomospecies of pathogenic Leptospira have been reported in New Caledonia (Salaün et al. 2006) permit a presumptive identification of the genomospecies based on melting temperature of the amplicon (Merien et al. 2005). Similarly, the MAT also frequently allows the identification of the infecting serovar (Levett 2001).

During this period, five patients died from PCR-confirmed leptospirosis (fatality rate: 3.7%), their mean age (60.3 years) being significantly higher than that of total cases. The total Potential Years of Life Lost (PYLL65, corresponding to the active age) was 38.6 years.

Of the 135 cases, 98 return forms were collected; 410 forms from negative patients were used as controls. The symptoms for negative cases are also reported in Table 1. The risk factors are summarized in Table 2: 77.6% of cases reported contact with animals (dogs, horses, cattle, pigs and others). Interestingly, 41% knew of the presence of rodents in or around their premises. Fourty-seven percentage declared bathing in rivers or streams, 26.5% fishing in freshwater and 21% were hunters. Though not included in the return form (therefore underestimated) walking barefoot was mentioned by 17.3% and walking in mud for 7.1%. Seven cases (7.1%) had occupations compatible with an occupational exposure and could therefore, based on local regulations, be considered as affected by an occupational leptospirosis.

Table 2.   Risk factors analysis
Exposure toAnimalsCattlePigsHorsesDogsRodents in/ around houseFreshwater fishingFreshwater swimmingHunting
% in 98 confirmed cases77.6%30%28.6%30.6%54.1%40.8%26.5%46.9%21.4%
% in 410 negative cases62.9%9.8%15.9%12.7%49.3%16.8%10.7%21.7%7.6%
Odds ratio (95% CI)2.04 (1.18–3.52)3.89 (2.18–6.92)2.12 (1.23–3.65)3.04 (1.75–5.26)1.21 (0.76–1.93)3.41 (2.05–5.66)3.00 (1.68–5.37)3.19 (1.96–5.19)3.33 (1.74–6.36)

The molecular diagnosis allowed the putative identification of 55 cases caused by L. interrogans (76%) and 17 caused by L. borgpetersenii (24%). Using MAT results, the serogroup involved was presumptively identified in 80 cases. Icterohaemorrhagiae was the most frequently observed (52%), Australis (20%) and Pyrogenes (15%) also contributing. These results are in good accordance with previous reports from New Caledonia (Merien & Perolat 1996; Bourée et al. 1999; Berlioz-Arthaud et al. 2007b). These serologically determined cases were tentatively assigned to one Leptospira genomospecies based on previous knowledge of the serovars present in New Caledonia (Merien & Perolat 1996; Salaün et al. 2006; Berlioz-Arthaud et al. 2007b). These results are summarized in Table 3a,b.

Table 3. Leptospira serovars (a) and genomospecies (b) identified as presumptively responsible for the 135 leptospirosis cases
Infecting serogroupNumber of casesPutatively assigned to
(a)
Ballum2Leptospira borgpetersenii
Tarassovi2
Unknown (PCR-diagnosed) or co-agglutinins70Diagnosed with qPCR Tm or based on strongest agglutinations
Australis14L. interrogans
Icterohaemorragiae or copenhagenii32
Panama1
Pomona1
Pyrogenes13
(b)
Leptospira interrogans
Identified (qPCR melting peaks)55114 (84.4%)
Presumed (serovar assignation)59
L. borgpetersenii
Identified (qPCR melting peaks)1721 (15.6%)
Presumed (serovar assignation)4

Of the 135 cases, 88 patients were hospitalized (65.2%), necessitating 47 ambulance trips. They remained at the hospital for an average 7.23 days (1–55 days), including 1.25 days (0–26) in an ICU, a total of 636 hospitalization days, including 110 days in ICU. Based on current fees, the direct medical cost of these 88 hospitalized patients was 622 894 €, diagnostic inclusive. The analysis and treatments for the remaining 47 patients not hospitalized totalized 14 486 €. The total number of lost workdays (restricted to patients aged 18–60-year-old) was estimated to 1431 days, an estimated 86 340 € social cost (based on guaranteed minimum wage). Altogether, the total cost for this 6-month and 135-case epidemic could be estimated to a total of 723 720 €, an average 5361 € per patient.

Of the cases for which the aetiological species was identified, L. interrogans (n = 55) appeared more frequently associated with severe forms, responsible for the five deaths and associated with an increased necessity of intensive care (0.89 day in ICU vs. 0.18 day for L. borgpetersenii cases, not significant because of high variability) though the global hospitalization duration was not different (5.25 vs. 5.06 days).

Discussion

Under the influence of the La Niña phenomenon, high rainfalls were associated with floods and an epidemic of human leptospirosis during the first semester 2008 in New Caledonia. This epidemic affected at least 135 people, was caused by several infecting strains and had a rural distribution, similar to previous outbreaks: the Bourail region still appears as a hotspot for leptospirosis (Merien & Perolat 1996; Bourée et al. 1999; Berlioz-Arthaud et al. 2007b) with an incidence of 500 new cases per 100 000 popuation over this 6-month period. Interestingly, the eastern coast village of Ponerihouen (Figure 1) appears to have a similarly high incidence (476/100 000) not reported in the earlier years (Berlioz-Arthaud et al. 2007b). The clinical presentation was similar to other descriptions (Levett 2001; Bharti et al. 2003) and the well-known Weil syndrome was present as a whole in less than one-third of the cases. This figure, though already reported, highlights the importance of medical awareness when facing flu-like syndromes, especially in tropical places where differential diagnosis is harder. Thanks to the awareness of leptospirosis in New Caledonia, most practitioners requested confirmation using PCR. The molecular detection of leptospiral DNA in blood or urine actually confirmed as much as 54% of the cases and allowed an earlier diagnosis than a sero-conversion would. The fatality rate of this epidemic was 3.7%, in the low range of classically reported rates (Smythe 1999) and was possibly related to the early diagnosis allowed by the PCR (Galloway et al. 2009). However, all five fatal cases were confirmed by PCR using sera (including one post-mortem diagnosis) and a delayed diagnosis does probably not account for four of these five fatal outcomes. Of these four fatal cases, three probably illustrate the fulminant forms of leptospirosis, and one was related to an underlying debilitating chronic pathology. Therefore, this low fatality rate possibly reflects the basal and inevitable fatality rate in clinical leptospirosis outbreaks, also related to the good awareness of clinicians, thereby limiting the underestimation of this disease.

These points highlight the usefulness of PCR as a first-line diagnosis of acute leptospirosis, possibly contributing to a relatively low PYLL65 when compared with other severe infections. In a 17-year retrospective evaluation (1991–2007) leptospirosis was directly responsible for 0.67% of the total PYLL65, much lower than accidents and traumatisms but accounting for 13.7% of infectious death causes (S. Laumond-Barny and J.P. Grangeon, unpublished data).

Using retrospective investigations, we estimated the cost of the 2008 epidemic to 5361 € per patient. Including the negative diagnostic costs, prevention and control costs, and for a total 157 cases during the 2008 year, the extrapolated 2008 total cost of the leptospirosis epidemic in New Caledonia is estimated at 984 472 € including costs of negative diagnosis, worth 3.95 € per capita based on 2008 estimates of the total population.

During the same period, an awareness programme (flyers, posters and a TV advert) and rodents control measures (restricted to the main city: Nouméa) had cost of 25 027 € and 27 050 €, respectively, a total prevention cost of 52 077 €. Altogether, the cost of leptospirosis and its control in New Caledonia in 2008 can be estimated as 4.16 € per capita, a non-negligible cost for a non-negligible disease.

Acknowledgements

The authors would like to acknowledge people in charge of the leptospirosis diagnostic at IPNC, namely L. Massenet, C. Manauté, D. Girault, D. Moleana, D. Triballi and S. Troc under the authority of I. Lecuyer, Drs Martine Noël and Anne Pfannstiel from the New Caledonian Health Authority. Thanks are also due to Meteo France for kindly providing meteorological records and to Dr Odile Qaeze for providing the cost of rodents control in Nouméa.

Ancillary