Población de referencia y muestra del estudio

G.Morroy H.H.J.Bor J.Polder J.L.A.Hautvast W.van der Hoek P.M.Schneeberger C.J.Wijkmans

European Journal of Public Health.2012; 22(6) 814–819 DOI: 10.1093/eurpub/cks003

114 | Chapter 6

ABSTRACT

In the Netherlands, 1,168 Q-fever pa ents were no fi ed in 2007 and 2008. Pa ents and general prac oners (GPs) regularly reported persis ng symptoms a er acute Q-fever, especially fa gue and long periods of sick leave, to the public health authori es. Inter- na onal studies on smaller Q-fever outbreaks demonstrate that symptoms may persist years a er acute illness. Data for the Dutch outbreaks were unavailable. The aim of this study is to quan fy sick leave and long-term symptoms a er acute Q-fever.

Methods

Our study targeted 898 acute Q-fever pa ents, no fi ed in 2007 and 2008 residing in the Province of Noord-Brabant. Pa ents from the 2008 cohort were mailed a ques onnaire at 12 months and those of the 2007 cohort at 12-26 months a er onset of illness. Pa ents reported underlying illness, Q-fever related symptoms and sick leave.

Results

The response rate was 64%. Forty percent of working pa ents reported long-term (>1 month) sick leave. Pre-existent heart disease odds ra o (OR) 4.50; confi dence interval (CI) 1.27-16.09), hospitalisa on in the acute phase (OR 3.99;95% CI 2.15-7.43) and smoking (OR 1.69; 95% CI 1.01-2.84) were signifi cant predictors for long-term absence. Of the pa ents who resumed work, 9% were, at the me of comple ng the ques on- naire, s ll unable to func on at pre-infec on levels due to fa gue or concentra on problems. Of the respondents 40% reported persis ng physical symptoms at the me of follow-up. Fa gue (20%) was most frequently reported. Daily ac vi es were aff ected in 30% of cases.

Conclusions

Self-reported sick leave and long-term health symptoms of Q-fever pa ents | 115

6

INTRODUCTION

Q-fever is a worldwide zoono c disease caused by Coxiella burne i (C. burne i), an obligate intracellular bacterium. In the Netherlands, Q-fever was uncommon before 2007 with 10-20

no fi ed cases annually [1].Since 2007 and up to December 2010 more than 4.000 cases

[2] were no fi ed in four major outbreaks implica ng dairy goats as the source [1,3-5].Ap-

proximately 80% of the no fi ed Q-fever pa ents reside in Noord-Brabant, the province with the highest dairy goat density in the Netherlands.

C. burne i is common in a wide range of wild and domes c animals but only small ru-

minants, in par cular sheep and goats, are associated with large human outbreaks [6,7].In-

fected animals excrete billions of bacteria in birth products and to a lesser extent in faeces, urine and milk. Human infec on occurs mainly a er inhaling dust par cles contaminated with C. burne i [7].

In suscep ble individuals infec on develops a er a mean incuba on period of 21 days [6]. In general, 60% of the infected Q-fever pa ents are asymptoma c, whereas 20% of the pa-

ents develop mild symptoms. Another 20%, however, present with more severe symptoms ranging from high fever, severe headache, night swea ng, nausea, diarrhoea, to pneumonia, hepa s, pericardi s, myocardi s and neurological symptoms [8].

Chronic Q-fever may develop in up to 5% of acute cases [9-11], due to a reac va on of Coxiella. A follow-up of 686 Dutch acute Q-fever pa ents from 2007 to 2008 found that 1.6% converted to chronic Q-fever [12]. A feared presenta on of chronic Q-fever is endocardi s [11] that may take 10-15 years to develop. Pregnant women and people with heart valve disorders, vascular prosthesis, or impaired immunity have a higher risk to develop a chronic infec on.

Q-fever pa ents may develop Q-fever fa gue syndrome (QFS), a debilita ng fa gue out

of propor on with exer onthat may last up to 10 years [13,14].Post-infec on fa gue [15] is

not unique for Q-fever. It may also occur a er other infec ous diseases such as Lyme disease, Epstein–Barr virus infec on [16], legionnaires disease [17], and other pneumonias [18].

Anecdotal informa on suggests that pa ents from the 2007 cohort had a more severe course of illness compared with those from the 2008 outbreak and were longer absent from work. However, evidence on the recovery of pa ents from the Dutch 2007 and 2008 cohorts is lacking. This study aims to fi ll that gap. The fi rst objec ve is to assess the dura on of sick leave a er an episode of acute Q-fever-in 2007 or 2008-and the long-term self-reported symptoms and associated risk factors. The second objec ve is to asses diff erences between the two cohorts for the dura on of self reported sick leave and the occurrence and frequency of long-term health symptoms.

116 | Chapter 6

METHODS

Study design and popula on

The popula on for this cohort study consisted of 898 pa ents no fi ed in 2007 and 2008 to the Municipal Health Services ‘Hart voor Brabant’ and ‘Brabant Zuid-Oost’. Due to incomplete data or an unknown month of onset of illness, 28 pa ents were excluded (Figure 1). The re- maining 870 Q-fever pa ents fi ed the Dutch no fi ca on criteria; a laboratory confi rma on of Q-fever and clinical presenta on with fever, pneumonia or hepa s

Data collec on

In February 2009, all pa ents received a study informa on folder including a par cipa on request and consent form by post (Figure 1). Pa ents could state their willingness to take part in any of a number of studies under the so-called Q Quest-1 project, by signing and returning the consent form. Par cipa ng pa ents received a ques onnaire -by postal mail -focussing on demographic characteris cs, medical history, smoking behaviour, current Q-fever related physi- cal symptoms and employment related items, such as dura on of sick leave following acute Q-fever, resump on and current ability to work (employed, self-employed, volunteer work, household work). During a pre-test, the comple on of the ques onnaire took 20 - 30 minutes.

All pa ents from the 2007 cohort received the ques onnaire in February 2009 (13-26 months a er onset of illness). Pa ents no fi ed in 2008 were mailed a ques onnaire one year a er the month of onset of illness. Pa ents from both cohorts, received two reminders a er three and six weeks. We obtained data on non-responders regarding gender, age, year of onset of disease and hospitaliza on from the no fi ca on data of the Municipal Health Services.

Data analysis

Ques onnaires were double scanned and data were cleaned. Sta s cal analysis was done using SPSS 16 for windows. We used the Mantel–Haenzel Chi square and Fishers exact test to compare propor ons. P-values are based on two tailed tests with a p-value<0.05 defi ned as signifi cant. The no fi ca on data of the Municipal Health Services were used to compare responders and non-responders for age, gender, year of onset of illness and hospitalisa on.

Mul variate logis c regression was used to model the rela onship between outcome (sick leave >1 month or presence of symptoms) and the independent variables age, gender, hospitalisa on, underlying diseases and year of onset of illness. Mul variate logis c regres- sion was used to model the rela onship between the determinants year of onset of illness and hospitalisa on with the outcomes presence of symptoms (fa gue)and long-term sick leave. For the poten al confounders, age, gender, smoking and co-morbidity , we used the same model.

Self-reported sick leave and long-term health symptoms of Q-fever pa ents | 117

6

RESULTS

Pa ent par cipa on

The overall response rate was 63.9% (Figure 1). The mean me between the day of onset of illness and receiving the ques onnaire was 19.5 months (SD 2.3) for pa ents of the 2007 cohort and 12.1 months (SD 0.5) for pa ents of the 2008 cohort. We were informed that fi ve pa ents died, but had no informa on on the cause of death.

The response rate of men was lower than that of women as was that of younger (≤30 years of age) compared with older pa ents (> 30 years of age). There were no diff erences between responders and non-responders for year of onset of illness and hospitaliza on (see Supplementary Table S1 for details).

Figure 1.Response rate of pa ents from cohorts 2007 to 2008. Division in cohorts is on the basis of

118 | Chapter 6

Characteris cs of the study popula on

The study popula on (Table 1) of the 2007 and 2008 cohorts was similar with respect to gender, age group, smoking behaviour and underlying diseases. The hospitalisa on rate of pa ents of the 2007 cohort, however, was signifi cantly higher [rela ve risk (RR) 2.50, 95% CI 1.68-3.01;p< 0.000] than that of the 2008 cohort. Pa ents from the 2007 cohort were more o en depressed (OR 1.88, 95% CI 1.10-3.34; p= 0.044) than pa ents from the 2008 cohort.

Sick leave

Prior to the Q-fever infec on 62% of the study popula on was gainfully employed. During the episode of acute fever 91.3% of these pa ents reported sick for work (Figure 2). Overall 132 (39.6%) of study subjects that worked prior to the infec on, were longer than one month (long- term sick leave) absent from work following an acute Q-fever infec on. See Supplementary Table S2 for sick leave of the gainfully employed, volunteers and those who do household work. Table 1. Characteris cs of the study popula on (n= 556) of two cohorts (2007 and 2008) of acute

Q-fever pa ents no fi ed in The Netherlands.

Characteris cs Cohort 2007, n = 96 (%) 2008, n = 460 Total, n = 556 p-value

Gender

Male 55 (57.3) 278 (60.4) 333 (60.0)

Female 41 (42.7) 182 (39.6) 223 (40.0) 0.568

Mean age in years (SD) 50.4 (14.9) 51.8 (13.3)

Smoker Yes 36 (39.2) 150 (33.6) 186 (34.6) No 56 (60.8) 296 (66.4) 352 (65.4) 0.317 Hospitalised Yes 42 (43.8) 89 (19.5) 131 (23.7) No 54 (56.2) 368 (80.5) 422 (76.3) <0.000 Underlying diseasea Yes 57 (59.4) 261 (56.7) 318 (57.2) No 39 (40.6) 199 (43.3) 238 (42.8) 0.639 Underlying diseases Diabetes 7 (7.3) 29 (6.3) 36 (6.5) 0.721 Heart disease 8 (8.3) 38 (8.3) 46 (8.3) 0.981 Lung disease 10 (10.4) 34 (7.4) 44 (7.9) 0.318 Arthri s 2 (2.1) 20 (4.3) 22 (3.9) 0.301 Depression 9 (9.4) 20 (4.3) 29 (5.2) 0.044 Other specifi ed 4 (4.2) 40 (8.7) 44 (7.9) 0.158 Other unspecifi ed 21 (21.9) 73 (15.9) 94 (16.9) 0.135 a

Self-reported sick leave and long-term health symptoms of Q-fever pa ents | 119

6