JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Journal of Infection and Public Health (2014) xxx, xxx—xxx
Defining travel-associated cases of enteric fever Joanne Freedman a,∗, Lorraine Lighton b, Jane Jones a a
Travel and Migrant Health Section, Public Health England Centre for Infectious Disease Surveillance and Control, 61 Colindale Avenue, London NW9 5EQ, UK b Greater Manchester Public Health England Centre, Health Protection Team, Floor 7B Sentinel House, Albert Street, Eccles, Manchester M30 0NJ, UK Received 20 March 2013 ; received in revised form 11 July 2013; accepted 29 July 2013
KEYWORDS Typhoid fever; Paratyphoid fever; Enteric fever; Foreign travel; Surveillance
Summary There is no internationally recognized case-definition for travelassociated enteric fever in non-endemic countries. This study describes the patterns of case reporting between 2007 and 2011 as travel-associated or not from the surveillance data in England, Wales and Northern Ireland (EWNI), before and after a change in the time component of the case-definition in January 2011. It examines in particular the role of a time frame based on the reported typical incubation period in defining a case of travel-associated enteric fever. The results showed no significant differences in the distribution of cases of enteric fever in regards to the interval between the onset and UK arrival in 2011 compared to 2007—2010 (p = 0.98 for typhoid and paratyphoid A); the distribution for paratyphoid B was also similar in both time periods. During 2007—2010, 93% (1730/1853) of all of the cases were classified as travel-associated compared to 94% (448/477) in 2011. This difference was not statistically significant. Changing the time component of the definition of travel-associated enteric fever did not make a significant difference to the proportion of travel-associated cases reported by investigators. Our analysis suggests that time might be subordinate to other considerations when investigators classify a case as travel-associated. Crown Copyright © 2014 Published by Elsevier Limited on behalf of King Saud Bin Abdulaziz University for Health Sciences. All rights reserved.
Introduction Enteric fevers (otherwise known as typhoid and paratyphoid fevers, which in this study are referred to as ‘typhoid’ and ‘paratyphoid’) are exclusively ∗
Corresponding author. Tel.: +44 020 8327 6412. E-mail address: [email protected] (J. Freedman).
human diseases transmitted via the fecal—oral route and caused by Salmonella enterica subspecies enterica serovar Typhi and Paratyphi A, B, and C. Unlike other Salmonella spp., they primarily cause a systemic illness with fever, headache, stomach pain, loss of appetite and nausea; diarrhea might be present in some cases, but not all. The illness varies in severity; paratyphoid has been reported in the literature to be typically
1876-0341/$ — see front matter. Crown Copyright © 2014 Published by Elsevier Limited on behalf of King Saud Bin Abdulaziz University for Health Sciences. All rights reserved.
http://dx.doi.org/10.1016/j.jiph.2013.07.011
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
2
J. Freedman et al.
milder than typhoid, although more recent studies have suggested that there is little difference between their clinical presentations [1,2], and both could be serious and life-threatening unless treated promptly with antibiotics. The incubation periods for enteric fevers are dependent on the infectious dose and the vehicle in which the organisms are ingested as well as on host factors [3,4]. Typical ranges quoted in the literature are 8—14 days and 10—20 days for typhoid and 1—10 days for paratyphoid [5] although the range might be wider, and the extreme range for typhoid is quoted as 3—60 days [5,6]. Enteric fevers occur in parts of the world where hygiene and sanitation are poor. The latest estimates for the number of typhoid and paratyphoid cases occurring globally were 21.7 million and 5.4 million, respectively, in 2000 [7]. The regions of high incidence are Africa, parts of South and South-East Asia, South America and several Asian regions of the former USSR [8]. Enteric fever in the United Kingdom [2] and other countries in western Europe [9—12], the United States [13], Canada [14], Australia [15], New Zealand [16], Japan [17] and Singapore [18] is most often associated with travel to or arrival from an endemic country. In many published studies on travel-associated enteric fever, ‘‘travel-associated’’ is not explicitly defined, or where it is defined, the time interval between travel and the onset of symptoms used in the definition differs across countries, with the maximum varying between three weeks and three months (Table 1). National enhanced surveillance of enteric fever in England, Wales and Northern Ireland (EWNI) began in May 2006, and all of the cases in which Salmonella Typhi or Paratyphi has been isolated from a blood (optimally) or fecal sample are eligible for inclusion. All of the isolates are confirmed and typed by the Salmonella Reference Service at Public Health England, Colindale. The protocol and questionnaire for the surveillance were developed by a multidisciplinary collaborative steering group drawn from laboratory and epidemiological specialists and practitioners in Public Health England (PHE) (formerly the Health Protection Agency), the Local Authorities Coordinators of Regulatory Services and the National Travel Health Network and Centre (NaTHNaC). The questionnaires are completed by Local Authority environmental health officers or the local PHE staff, depending on the local arrangements, and coordinated nationally by the Travel and Migrant Health Section (TMHS) at the PHE Centre for Infectious Disease Surveillance and Control in Colindale. Between 2007 and 2011, questionnaires were
submitted to TMHS for 86% of laboratory-confirmed cases. Until 2011, a travel-associated case of enteric fever was defined in the national surveillance questionnaire as onset within three weeks of arrival in or return to the UK from abroad. This definition was based on the typical maximum incubation period quoted for typhoid [25] and expert consensus from the steering group. For simplicity, the identical definition was used for typhoid and paratyphoid despite their reportedly different incubation periods because, at the time of completion of the questionnaire, the causative organism might not be known. Analysis of the data collected between 2007 and 2010 demonstrated that investigators did not always adhere to the three-week cut off period for defining a travel-associated case of enteric fever, particularly in cases in which there had been travel to a known endemic country outside the defined time frame. The vast majority of cases between 2007 and 2010 had an onset of illness up to one month after arrival in or return to the UK from abroad. In the light of these findings, the definition of a travel-associated case was changed in January 2011 to cases that had an onset of illness within one month after arrival in or return to the UK from abroad. For analysis purposes, a month was defined as 28 days. This study retrospectively describes the patterns of reporting of cases as travel-associated or not, before and after the change in definition of a travel-associated case, to examine how much weight investigators gave to the time frame of the definition and to consider the appropriateness of the change in definition. The general epidemiology of enteric fevers in EWNI is beyond the scope of this study and is available elsewhere [2].
Methods The data from national surveillance are managed with a Microsoft Access 2007 database and were analyzed for this study using Microsoft Excel 2007. In cases in which a questionnaire could not be completed, information on the travel history might be obtained from the laboratory request form (for the country of travel only) or directly from the local PHE Health Protection Teams. The investigators completing questionnaires are asked to classify a case as travel-associated or not (based on the definitions as described above) and to complete information about the onset date of illness and dates of any travel. The time (in days) between overseas travel and the onset date was calculated for all of the
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever Table 1
3
Definitions of travel-associated enteric fever in developed countries.
Country
Definition of travel-associated enteric fever
Geographical element
Incubation period included in definition
Time element
Canada [14]
If traveled to an endemic area during at least part of the incubation period (21 days for typhoid, 10 days for paratyphoid) or had traveled to a non-endemic area outside of Canada for the entire incubation period. For a travel-associated case: the delay between the departure and onset must be greater than or equal to the minimum incubation period (7 days) and the delay between the return and onset dates must be less than the maximum incubation period (21 days). Cases occur within a month of return from a stay abroad. Cases of enteric fever occurring within 6 weeks of recent travel. Culture confirmed or serologically confirmed acute typhoid fever with compatible symptoms in a person who has been in a developing country for a 4-week period preceding onset of illness. First day of illness at least 3 days after departure and within 1 month of return from (sub) tropical areas. Cases that report overseas travel during the incubation period (8—14 days typhoid; 1—10 days paratyphoid stated in Communicable Diseases Manual [22]). Cases were classified as imported if there was a history of travel to an endemic country between 7 and 21 days before the onset of symptoms. Non-travel cases were those in which the case had not traveled within 3 months before the onset. Traveled outside the US in the 6 weeks before illness onset. Travel-associated typhoid fever is defined as a case of typhoid fever meeting the national surveillance case definition in a person who traveled outside of the United States in the 30 days before illness.
All countries
Yes
21 days typhoid, 10 days paratyphoid
All countries
Yes
Between 7 and 21 days
All countries
No
1 month
All countries
No
6 weeks
Developing country
No
4 weeks
(Sub) tropical areas
No
Between 3 days and 1 month
All countries
Yes
1—3 weeks
Endemic area
No
Between 7 and 21 days
All countries
No
3 months
All countries
No
6 weeks
All countries
No
30 days
Canada [19]
France [10] Israel [20] Netherlands [21]
Netherlands [12]
New Zealand [16]
Singapore [23]
Switzerland [11]
USA [24] USA [13]
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
4
J. Freedman et al.
symptomatic cases reported through the enhanced surveillance scheme between 2007 and 2011 inclusive, where both dates were known. For the purpose of this study, the time between overseas travel and the onset date is defined from the date of return to or arrival in the UK and includes those whose trip originated from the UK and those who were new entrants or visitors to the UK. Cases were excluded from the analysis if they: had more than one organism isolated (at the species level only; different phage and antibiotic resistance types were not considered in this study), had no onset or travel dates, had onset before departure from the UK, or had onset before the minimum quoted incubation period after arrival in an endemic country (one day for paratyphoid and three days for typhoid). Cases reported between 2007 and 2010 were compared with those reported in 2011 in terms of the time between the onset date and overseas travel and the correspondence between the investigator classification and the operational travel-associated definition. The Mann—Whitney U test was calculated at the 5% level (twotailed) using IBM SPSS Statistics Version 19 to assess whether there were any differences in the non-normally distributed time intervals between the onset date and the dates of overseas travel between cases of typhoid and paratyphoid A during 2007—2010 and between 2007—2010 and 2011. The two-independent samples z test at the 95% level was calculated using STATA (STATA/SE 12.0 for Windows, Stata Corp LP., USA) to assess the significance of the difference between the proportions of cases classified as travel-associated in the two time periods. It is assumed that all of the investigators administered the questionnaires identically and that all of the cases had an equal chance of being interviewed and of answering all of the questions.
Table 2
Results Completeness of reporting Table 2 shows the completeness of reporting for the onset and travel history of cases during 2007—2010 and 2011. Reporting was much improved in 2011 with 83.2% of travel-associated cases, as designated by the investigators, having travel dates and onset date, compared to 65.2% during 2007—2010.
Cases reported between 2007 and 2010 Between 2007 and 2010, 1966 laboratory-confirmed symptomatic cases of typhoid and paratyphoid A and B (single isolates only) were reported in EWNI (an average of 492 each year). Of the cases with a complete travel history (N = 1845), 1725 (93.5%) were classified as travel-associated by the investigator, and 123 (6.5%) were classified as nontravel-associated. The non-travel-associated cases comprised the following: 37 specifically stated they had not traveled at all; 24 had traveled to an endemic country but the onset was before departure from the UK or occurred more than 21 days after overseas travel; and for the remaining 62 cases, it was unknown whether there had been no travel or if there had been travel outside of the incubation period. This information was not specifically requested on the questionnaire. Between 2007 and 2010, after the exclusion criteria were applied, a total of 1274 cases of travel-associated enteric fever (678 cases of typhoid, 549 cases of paratyphoid A and 47 cases of paratyphoid B) were included for analysis of time intervals between overseas travel and onset; 24% (297/1227) of typhoid and paratyphoid A cases became ill while they were abroad (Fig. 1). The majority of the remaining cases (460/511, 90.0%
Completeness of reporting travel and onset dates for enteric fever cases: 2007—2011.
Variables
Total symptomatic cases Travel history of symptomatic cases as designated by investigator Travel-associated cases designated by investigator Travel-associated cases designated by investigator: With travel dates With onset date With both travel and onset dates With both travel and onset dates and excluding cases with onset before travel
N cases (% total) 2007—2010
2011
1966 1845 (93.5)
481 477 (99.2)
1725 (87.4)
448 (93.1)
1262 (64.2) 1324 (67.3) 1281 (65.2) 1274 (64.8)
374 (77.8) 408 (84.8) 400 (83.2) 394 (81.9)
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever
5
Figure 1 Time (days) between onset date and arrival in the UK by organism with cumulative proportions for the cases designated as travel-associated (N = 1274, 65% of the total cases): 2007—2010.
typhoid; 359/419, 85.7% paratyphoid A) became ill within 21 days after overseas travel and a further 64 (25/511, 4.9% typhoid and 39/419, 9.3% paratyphoid A) became ill between 22 and 28 days after overseas travel, with a median of eight days for typhoid and nine days for paratyphoid A. There was no statistically significant difference in the distribution of time intervals between onset and overseas travel between typhoid and paratyphoid A (p = 0.12). For paratyphoid B, 43% (20/47) of cases became ill while they were abroad, and the remainder (27/47, 57%) became ill within 21 days after overseas travel, most (19/27, 70%) within the first seven days (median four days).
Cases reported in 2011 In 2011, 481 laboratory-confirmed symptomatic cases of typhoid and paratyphoid A and B (single isolates only) were reported in EWNI. Of cases with a completed travel history (477), 448 (93.9%) were designated as travel-associated by the investigator, and 29 (6.1%) were designated as nontravel-associated. The non-travel-associated cases comprised the following: nine cases specifically stated they had not traveled at all; one had traveled to an endemic country, though not recently (no dates were given); for the remaining 19 cases, it was unknown whether there had been no travel, or if there had been travel outside of the incubation period. When the exclusion criteria were applied, a total of 394 travel-associated cases of enteric fever
(203 cases of typhoid, 184 cases of paratyphoid A and seven cases of paratyphoid B) were included for analysis of time between onset and overseas travel; 21% (80/387) of the typhoid and paratyphoid A cases became ill while they were abroad, whereas 43% (3/7) of the paratyphoid B cases became ill while they were abroad (Fig. 2). The majority of the remaining cases (151/159 typhoid 95%; 141/148 paratyphoid A 95%; 3/7 paratyphoid B 43%) became ill within 28 days of overseas travel, with a median of seven days for typhoid and nine days for paratyphoid A. There was no statistically significant difference in the distribution of the time intervals between onset and overseas travel between typhoid and paratyphoid A (p = 0.41). There were no significant differences in the distribution of cases of typhoid and paratyphoid A in regards to the interval between onset and overseas travel in 2011 compared to 2007—2010 (p = 0.98 for typhoid and paratyphoid A); the distribution for paratyphoid B was similar in 2011, except for one case of paratyphoid B with onset of symptoms 56 days after return to the UK (Fig. 2).
Classification of travel-associated and non-travel-associated cases before and after the change in definition There was no significant difference in the proportion of total cases classified as travel-associated and non-travel-associated after the definition of a travel-associated case changed from those with
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
6
J. Freedman et al.
Figure 2 Time (days) between the onset date and arrival in the UK by organism with cumulative proportions for the cases designated as travel-associated (N = 394, 82% of total cases): 2011.
onset within 28 days compared to those within 21 days of overseas travel (Table 3). If the case definitions in use for each time period had been strictly applied to designate the travel status of the case, then the number of potentially misclassified UK acquired cases would have increased by 111 (90.2%) in 2007—2010 and by 16 (55.2%) in 2011. It is unclear from the information given on the surveillance forms whether further investigations were undertaken to seek a potential UK source in the cases that did not strictly fit the case definition; however, all but two of these cases were known to have recently arrived from or had traveled to a known endemic country, most on the Indian subcontinent. One case classified as travelassociated outside the case definition in 2008 had traveled to Germany, a country not typically known to be endemic for enteric fever as classified by NaTHNaC, and one case in 2007 had no country of travel stated.
Discussion There are a number of challenges associated with surveillance of travel-associated infections in developed countries [26], in particular the ascertainment of complete travel and clinical histories and the absence of an international ‘gold standard’ case-definition for travel- and non-travelassociated cases. Enhanced surveillance has greatly improved the completeness and quality of the data collection on enteric fever in these respects and has enabled a more accurate assessment of the enteric fever situation in EWNI [2]. Although enteric fever is primarily a travel-associated infection in the UK, sporadic cases associated with secondary transmission from a travel-associated case [28] and small clusters that are presumed to have been acquired in the UK have infrequently occurred [29—31]. More rigorous and time-consuming investigation and public health action is required for
Table 3 Designation of enteric fever as travel-associated or not by investigators, pre and post the change in definition of a travel-associated case: 2007—2011. Designation of cases by investigator
2007—2010
2011
% Change between 2007—2010 and 2011
Travel-associated cases (% of total)
1725 (93.5) (95% CI: 92.4—94.6) 123 (6.5) (95% CI: 5.4—7.6) 1845
448 (93.9) (95% CI: 91.8—96.0) 29 (6.1) (95% CI: 4.0—8.2) 477
Increased by 0.4% — not significant (p = 0.75)
Non travel-associated cases (% of total)
Total with known travel history
Decreased by 0.4% — not significant (p = 0.75)
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever non-travel-associated cases; therefore, it is important that cases be classified appropriately. One means of appropriate classification has been on the basis of a temporal relationship between symptom onset and travel. The initial definition of a travel-associated case as symptom onset within a three-week (21 day) period of overseas travel was not always adhered to by investigators. Differing incubation periods stated in various UK guidelines for investigating enteric fever in use at the time [25,32] might partly account for the differing assessments, but not strictly adhering to the initial definition might also reflect the fact that ‘time since travel’ is only one factor that investigators consider in assessing cases (for example, 4% of cases in 2007—2010 and 2011 were designated as travel-associated when they had onset after 28 days from overseas travel). Although there are a number of other questions included in the surveillance questionnaire [2] to ascertain whether a case might have been acquired in the UK, the reason for the designation as travel-associated or not is not always clear. This study shows that changing the definition of travel-associated cases of enteric fever in terms of the time between symptom onset and overseas travel did not significantly affect the proportion of cases designated by investigators as travel-associated. Our analysis suggests that time might be subordinate to other considerations when investigators classify a case, if, for example, there is no other plausible source of infection for the individual other than likely exposure abroad, especially in cases in which travel to an endemic country is known. This finding is consistent with a Norwegian study that considered the influence of time between return to Norway and symptom onset for other gastrointestinal (GI) infections; the study concluded that clinicians are likely to take into account the usual epidemiology of the infection within the home country when assessing whether a GI infection is travel-associated even if there is a long period between return to the home country and symptom onset [33]. A high proportion of typhoid and paratyphoid fever cases in EWNI have traveled abroad to visit friends and relatives [27], and many travel together in family groups; therefore, it is possible that secondary cases within a traveling group might be mistaken for primary travel-associated cases, which may account for the longer apparent incubation period in some cases. The purpose of a time frame is to provide guidance to investigators on which cases are very likely to be travel-associated and to help identify the cases in which further consideration should be given to the likelihood of them being UK-acquired (i.e., to prompt the investigation of other possible
7 risk factors), which is of significance because UK acquired cases require higher levels of public health action. One of the concerns that might be raised about extending the time frame for defining a case as travel-associated is that non-travel-associated cases might be consequently misclassified and not appropriately investigated and managed, leading to clusters of indigenous disease. There is no evidence from the enhanced surveillance that the change in definition has led to more UK-acquired cases of enteric fever being misclassified in 2011 than in previous years. The limitations of this study predominantly relate to the completeness of data capture, although this aspect has improved over time. In particular, for most cases classified as non-travelassociated, the basis on which this classification was made was not clear (e.g., because of the temporal relationship between travel and symptoms or because there had been no travel); this information is not requested on the form, but it might be worth revising the questionnaire to include this information for a future audit. Limited information was provided on the surveillance form about possible sources for non-travel-associated infections. Some sources are likely to be as a result of contact with known cases within a household; between 2007 and 2011, 14% (22/152) of non-travel-associated cases were likely to have been infected from a household contact who might or might not have traveled themselves. Subsequent to this study, and in part based on its findings, new ‘Public Health Operational Guidelines for Enteric Fever (Typhoid and Paratyphoid)’ [34] were published in early 2012. These guidelines aim to improve the investigation of nontravel-associated cases, providing algorithms for investigators to use in trying to identify potential sources of infection in the absence of a travel history within 28 days of the onset of symptoms. Part of the evaluation of these guidelines [35] includes a detailed analysis of cases known or suspected to be acquired in the UK to assess the value of more thorough investigation of cases with a long period of time between overseas travel and the onset date. It is anticipated that data completeness will be much improved subsequent to the implementation of these guidelines. For those deciding the public health actions required in response to a case of enteric fever, the temporal definition provides useful guidance (in conjunction with other information about the case) as to what public health actions are required. It might be that once the investigation of the case is complete and also considering the possibility of any contact with co-travelers, travel abroad is the most likely source in the absence of evidence to the
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
8
J. Freedman et al.
contrary, even if the case had onset of symptoms more than 28 days after arrival in or return to the UK. For surveillance reporting and data analysis purposes, it is appropriate to continue to rely on the professional judgment of the public health practitioner to define a case as travel-associated (or presumed travel-associated) as is currently the case [2,28]. In this analysis, the vast majority of enteric fever cases reported in EWNI between 2007 and 2011, in which the individuals did not become ill while they were abroad, had onset dates within 28 days of arrival in or return to the UK, with a median of nine days and seven days for paratyphoid A and typhoid, respectively, and a median of five days for paratyphoid B. The data demonstrate that, although the median for typhoid was slightly lower, there was no significant difference between typhoid and paratyphoid A in this respect, suggesting similar incubation periods for these organisms; however, the data suggest that paratyphoid B has a shorter typical incubation period. The upper limit of the typical incubation period quoted in the literature for enteric fevers is lower than that observed in this case series. This difference is especially striking because this analysis measures time from arrival in the UK to onset as a proxy for the incubation period, rather than the time from infection to onset, so that the actual maximum incubation periods observed in this study are likely to be longer. For enteric fever, as well as many other infections, incubation periods vary between individuals, depending on the infecting dose and the immune status and general health of the individual who has been infected. Using the duration between travel and symptom onset is only one factor to consider in determining whether or not a case is travel-associated.
Author contributions Joanne Freedman was responsible for the concept and design of the study, the analysis and interpretation of the data, and the drafting of and final approval of the article. Lorraine Lighton was responsible for advising on the interpretation of the data, critical appraisal of the article for the intellectual content and final approval of the article. Jane Jones was responsible for advising on the concept and design of the study and data interpretation, critical appraisal of the article for the intellectual content and final approval of the article.
Conflicts of interest Funding: No funding sources. Competing interests: We have no conflict of interest to declare. Ethical approval: Not required.
Acknowledgments The authors gratefully acknowledge the following: Our health protection colleagues in local PHE Centres and the Environmental Health Officers who completed the questionnaires for the enhanced enteric fever surveillance; our colleagues in the PHE Salmonella Reference Service for providing the data on laboratory-confirmed cases of enteric fever for the enhanced surveillance; Edgar Wellington, Travel and Migrant Health Section, who managed the daily administration of the enteric fever enhanced surveillance; Helen Green from the PHE Respiratory Diseases Department for providing statistical advice.
Conclusion Epidemiology is a complex science, and although evidence-based guidelines for best practice are useful in investigating cases of infectious disease, public health professionals must use their professional judgment when a case does not appear to be consistent with the rules. The key is to be able to gather as much information as possible about the case for accurate and consistent reporting and analysis of the data. Surveillance will continue to provide evidence to inform and improve the public health management of cases and to improve the advice given to travelers who might be at increased risk of acquiring enteric fever through overseas travel.
References [1] Maskey AP, Day JN, Tuan PQ, Thwaites GE, Campbell JI, Zimmerman M, et al. Salmonella enterica Serovar Paratyphi A and S. enterica Serovar Typhi cause indistinguishable clinical syndromes in Kathmandu, Nepal. CID 2006;42(May (1)):1247—53. [2] Public Health England (PHE). Enhanced surveillance of enteric fever webpage. Available on the HPA legacy website http://www.hpa.org.uk/Topics/InfectiousDiseases/ at: InfectionsAZ/TravelHealth/GeneralInformation/trav30 Enhancedsurveillanceofentericfever/ [accessed 10.06.13]. [3] Glynn JR, Bradley DJ. The relationship between infecting dose and severity of disease in reported outbreaks of Salmonella infections. Epidemiol Infect 1992;109:371—88. [4] Miner JR. The incubation period of typhoid fever. J Infect Dis 1922;31:296—301.
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever
9
[5] Gillespie S. Salmonella Infections. In: Cook GC, Zumla A, editors. Manson’s tropical diseases. Edinburgh: Saunders; 2003. p. 937—49. [6] Mintz E, Sodha S. Typhoid fever. In: Heymann DL, editor. Control of communicable diseases manual. Washington: American Public Health Association; 2008. p. 664—71. [7] Crump JA, Luby SP, Mintz ED. The global burden of typhoid fever. Bull World Health Organ 2004;82:346—53. [8] World Health Organization. Diarrhoeal diseases: typhoid fever. Geneva World Health Organization Initiative for Vaccine Research; 2009. Available at http://www.who.int/ vaccine research/diseases/diarrhoeal/en/index7.html [last accessed 5.02.13]. [9] European Centre for Disease Prevention and Control (ECDC). Typhoid/paratyphoid fever. In: ECDC, editor. Annual epidemiological report 2011. Stockholm: ECDC; 2011. p. 116. [10] Delmas G, Vaillant V, Jourdan N, Le Hello S, Weill FX, de Valk H. Typhoid and paratyphoid fever in France between 2004 and 2009. Bull Epidemiol Hebd 2011;2:9—12. [11] Keller A, Frey M, Schmid H, Steffen R, Walker T, Schlagenhauf P. Imported typhoid fever in Switzerland, 1993 to 2004. J Travel Med 2008;15:248—51. [12] Bovee LP, van Kessel RP, Peerbooms PG, van den Hoek JA. Buiktyfus in Amsterdam, 1991—2000, en vereenvoudiging van het landelijk protocol voor bron- en contactopsporing [Typhoid fever in Amsterdam, 1991—2000, and the simplification of the national protocol for source and contact tracing]. Ned Tijdschr Geneeskd 2002;146: 1833—7. [13] Lynch MF, Blanton EM, Bulens S, Polynak C, Vojdani J, Stevenson J, et al. Typhoid fever in the United States, 1999—2006. JAMA 2009;302:859—65. [14] Taylor M, MacDougall L, Li M, Galanis E. The impact of international travel on the epidemiology of enteric infections, British Columbia, 2008. Can J Public Health 2010;101:332—6. [15] Skull SA, Tallis G. An evidence-based review of current guidelines for the public health control of typhoid in Australia: a case for simplification and resource savings. Aust N Z J Public Health 2001;25:539—42. [16] The Institute of Environmental Science and Research Ltd., Porirua, New Zealand Notifiable and other diseases in New Zealand: annual report 2010; 2011. [17] Shigematsu M, Okabe N. Trends of statutory reporting of travel associated typhoid fever, paratyphoid fever, cholera, and HAV infection in recent five years in Japan. In: Poster session presented at: 9th conference of the international society of travel medicine (CISTM9). 2005. [18] Ty AU, Ang GY, Ang LW, James L, Goh KT. Changing epidemiology of enteric fevers in Singapore. Ann Acad Med Singapore 2010;39:889—98. [19] Ravel A, Nesbitt A, Marshall B, Sittler N, Pollari F. Description and burden of travel-related cases caused by enteropathogens reported in a Canadian community. J Travel Med 2011;18:8—19.
[20] Meltzer E, Sadik C, Schwartz E. Enteric fever in Israeli travellers: a nationwide study. J Travel Med 2005;12:275—81. [21] Baaten GG, Sonder GJ, Van Der Loeff MF, Coutinho RA, van den HA. Fecal-orally transmitted diseases among travelers are decreasing due to better hygienic standards at travel destination. J Travel Med 2010;17:322—8. [22] Ministry of Health New Zealand. Communicable disease control manual 2012. Wellington: Ministry of Health; 2012. [23] Yew FS, Goh KT, Lim YS. Epidemiology of typhoid fever in Singapore. Epidemiol Infect 1993;110:63—70. [24] Ackers ML, Puhr ND, Tauxe RV, Mintz ED. Laboratory-based surveillance of Salmonella serotype Typhi infections in the United States: antimicrobial resistance on the rise. JAMA 2000;283:2668—73. [25] Hawker J, Begg N, Blair I, Reintjes R, Weinberg J. Communicable disease control handbook. 2nd ed. Oxford: Blackwell Science; 2005. [26] Lawrence J, Jones J, Hill DR. Improving the evidence base for pre-travel advice: the importance of surveillance of travel-associated infection. Br J Gen Pract 2005;55:566—8. [27] Health Protection Agency. Enteric fever epidemiological reports. Available at http://www.hpa.org.uk/web/ HPAweb&HPAwebStandard/HPAweb C/1259152344471 [28] Health Protection Agency. Enteric fever surveillance quarterly report (England, Wales and Northern Ireland): fourth quarter 2011. Health Protection Report 2012; 6: enteric. Available at http://www.hpa.org.uk/hpr/archives/2012/ hpr0612.pdf [29] Logan S, Jayasena C, Aali A, Ash S, Cooke FJ. Typhoid without travel. Clin Med 2010;10:299—300. [30] Public Health Laboratory Service. Typhoid fever outbreak in Newport, Wales: update. Commun Dis Rep CDR Wkly 2001;11(33), news. Available at http://www.hpa.org.uk/ cdr/archives/2001/cdr3301.pdf [31] Public Health Laboratory Service. Typhoid in North London. Commun Dis Rep CDR Wkly 1991;1(6), news. Available at http://www.hpa.org.uk/cdr/archives/1991/cdr0691.pdf [32] PHLS Advisory Committee on Gastrointestinal Pathogens. Preventing person-to-person spread following gastrointestinal infections: guidelines for public health physicians and environmental health officers. Commun Dis Public Health 2004;7:362—84. [33] Guzman-Herrador B, Vold L, Nygard K. Surveillance of travel-associated gastrointestinal infections in Norway, 2009—2010: are they all actually imported? Euro Surveill 2012;17(41), pii=20294. Available online at: http://www. eurosurveillance.org/ViewArticle.aspx?ArticleId=20294 [34] Health Protection Agency (HPA) and Chartered Institute of Environmental Health. Public health operational guidelines for enteric fever (typhoid and paratyphoid). London: HPA; 2012. [35] Balasegaram S, Potter AL, Grynszpan D, Barlow S, Behrens RH, Lighton L, et al. Guidelines for the public health management of typhoid and paratyphoid in England: practice guidelines from the National Typhoid and Paratyphoid Reference Group. J Infect 2012;65:197—213.
Available online at www.sciencedirect.com
ScienceDirect
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
ARTICLE IN PRESS
Journal of Infection and Public Health (2014) xxx, xxx—xxx
Defining travel-associated cases of enteric fever Joanne Freedman a,∗, Lorraine Lighton b, Jane Jones a a
Travel and Migrant Health Section, Public Health England Centre for Infectious Disease Surveillance and Control, 61 Colindale Avenue, London NW9 5EQ, UK b Greater Manchester Public Health England Centre, Health Protection Team, Floor 7B Sentinel House, Albert Street, Eccles, Manchester M30 0NJ, UK Received 20 March 2013 ; received in revised form 11 July 2013; accepted 29 July 2013
KEYWORDS Typhoid fever; Paratyphoid fever; Enteric fever; Foreign travel; Surveillance
Summary There is no internationally recognized case-definition for travelassociated enteric fever in non-endemic countries. This study describes the patterns of case reporting between 2007 and 2011 as travel-associated or not from the surveillance data in England, Wales and Northern Ireland (EWNI), before and after a change in the time component of the case-definition in January 2011. It examines in particular the role of a time frame based on the reported typical incubation period in defining a case of travel-associated enteric fever. The results showed no significant differences in the distribution of cases of enteric fever in regards to the interval between the onset and UK arrival in 2011 compared to 2007—2010 (p = 0.98 for typhoid and paratyphoid A); the distribution for paratyphoid B was also similar in both time periods. During 2007—2010, 93% (1730/1853) of all of the cases were classified as travel-associated compared to 94% (448/477) in 2011. This difference was not statistically significant. Changing the time component of the definition of travel-associated enteric fever did not make a significant difference to the proportion of travel-associated cases reported by investigators. Our analysis suggests that time might be subordinate to other considerations when investigators classify a case as travel-associated. Crown Copyright © 2014 Published by Elsevier Limited on behalf of King Saud Bin Abdulaziz University for Health Sciences. All rights reserved.
Introduction Enteric fevers (otherwise known as typhoid and paratyphoid fevers, which in this study are referred to as ‘typhoid’ and ‘paratyphoid’) are exclusively ∗
Corresponding author. Tel.: +44 020 8327 6412. E-mail address: [email protected] (J. Freedman).
human diseases transmitted via the fecal—oral route and caused by Salmonella enterica subspecies enterica serovar Typhi and Paratyphi A, B, and C. Unlike other Salmonella spp., they primarily cause a systemic illness with fever, headache, stomach pain, loss of appetite and nausea; diarrhea might be present in some cases, but not all. The illness varies in severity; paratyphoid has been reported in the literature to be typically
1876-0341/$ — see front matter. Crown Copyright © 2014 Published by Elsevier Limited on behalf of King Saud Bin Abdulaziz University for Health Sciences. All rights reserved.
http://dx.doi.org/10.1016/j.jiph.2013.07.011
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
2
J. Freedman et al.
milder than typhoid, although more recent studies have suggested that there is little difference between their clinical presentations [1,2], and both could be serious and life-threatening unless treated promptly with antibiotics. The incubation periods for enteric fevers are dependent on the infectious dose and the vehicle in which the organisms are ingested as well as on host factors [3,4]. Typical ranges quoted in the literature are 8—14 days and 10—20 days for typhoid and 1—10 days for paratyphoid [5] although the range might be wider, and the extreme range for typhoid is quoted as 3—60 days [5,6]. Enteric fevers occur in parts of the world where hygiene and sanitation are poor. The latest estimates for the number of typhoid and paratyphoid cases occurring globally were 21.7 million and 5.4 million, respectively, in 2000 [7]. The regions of high incidence are Africa, parts of South and South-East Asia, South America and several Asian regions of the former USSR [8]. Enteric fever in the United Kingdom [2] and other countries in western Europe [9—12], the United States [13], Canada [14], Australia [15], New Zealand [16], Japan [17] and Singapore [18] is most often associated with travel to or arrival from an endemic country. In many published studies on travel-associated enteric fever, ‘‘travel-associated’’ is not explicitly defined, or where it is defined, the time interval between travel and the onset of symptoms used in the definition differs across countries, with the maximum varying between three weeks and three months (Table 1). National enhanced surveillance of enteric fever in England, Wales and Northern Ireland (EWNI) began in May 2006, and all of the cases in which Salmonella Typhi or Paratyphi has been isolated from a blood (optimally) or fecal sample are eligible for inclusion. All of the isolates are confirmed and typed by the Salmonella Reference Service at Public Health England, Colindale. The protocol and questionnaire for the surveillance were developed by a multidisciplinary collaborative steering group drawn from laboratory and epidemiological specialists and practitioners in Public Health England (PHE) (formerly the Health Protection Agency), the Local Authorities Coordinators of Regulatory Services and the National Travel Health Network and Centre (NaTHNaC). The questionnaires are completed by Local Authority environmental health officers or the local PHE staff, depending on the local arrangements, and coordinated nationally by the Travel and Migrant Health Section (TMHS) at the PHE Centre for Infectious Disease Surveillance and Control in Colindale. Between 2007 and 2011, questionnaires were
submitted to TMHS for 86% of laboratory-confirmed cases. Until 2011, a travel-associated case of enteric fever was defined in the national surveillance questionnaire as onset within three weeks of arrival in or return to the UK from abroad. This definition was based on the typical maximum incubation period quoted for typhoid [25] and expert consensus from the steering group. For simplicity, the identical definition was used for typhoid and paratyphoid despite their reportedly different incubation periods because, at the time of completion of the questionnaire, the causative organism might not be known. Analysis of the data collected between 2007 and 2010 demonstrated that investigators did not always adhere to the three-week cut off period for defining a travel-associated case of enteric fever, particularly in cases in which there had been travel to a known endemic country outside the defined time frame. The vast majority of cases between 2007 and 2010 had an onset of illness up to one month after arrival in or return to the UK from abroad. In the light of these findings, the definition of a travel-associated case was changed in January 2011 to cases that had an onset of illness within one month after arrival in or return to the UK from abroad. For analysis purposes, a month was defined as 28 days. This study retrospectively describes the patterns of reporting of cases as travel-associated or not, before and after the change in definition of a travel-associated case, to examine how much weight investigators gave to the time frame of the definition and to consider the appropriateness of the change in definition. The general epidemiology of enteric fevers in EWNI is beyond the scope of this study and is available elsewhere [2].
Methods The data from national surveillance are managed with a Microsoft Access 2007 database and were analyzed for this study using Microsoft Excel 2007. In cases in which a questionnaire could not be completed, information on the travel history might be obtained from the laboratory request form (for the country of travel only) or directly from the local PHE Health Protection Teams. The investigators completing questionnaires are asked to classify a case as travel-associated or not (based on the definitions as described above) and to complete information about the onset date of illness and dates of any travel. The time (in days) between overseas travel and the onset date was calculated for all of the
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever Table 1
3
Definitions of travel-associated enteric fever in developed countries.
Country
Definition of travel-associated enteric fever
Geographical element
Incubation period included in definition
Time element
Canada [14]
If traveled to an endemic area during at least part of the incubation period (21 days for typhoid, 10 days for paratyphoid) or had traveled to a non-endemic area outside of Canada for the entire incubation period. For a travel-associated case: the delay between the departure and onset must be greater than or equal to the minimum incubation period (7 days) and the delay between the return and onset dates must be less than the maximum incubation period (21 days). Cases occur within a month of return from a stay abroad. Cases of enteric fever occurring within 6 weeks of recent travel. Culture confirmed or serologically confirmed acute typhoid fever with compatible symptoms in a person who has been in a developing country for a 4-week period preceding onset of illness. First day of illness at least 3 days after departure and within 1 month of return from (sub) tropical areas. Cases that report overseas travel during the incubation period (8—14 days typhoid; 1—10 days paratyphoid stated in Communicable Diseases Manual [22]). Cases were classified as imported if there was a history of travel to an endemic country between 7 and 21 days before the onset of symptoms. Non-travel cases were those in which the case had not traveled within 3 months before the onset. Traveled outside the US in the 6 weeks before illness onset. Travel-associated typhoid fever is defined as a case of typhoid fever meeting the national surveillance case definition in a person who traveled outside of the United States in the 30 days before illness.
All countries
Yes
21 days typhoid, 10 days paratyphoid
All countries
Yes
Between 7 and 21 days
All countries
No
1 month
All countries
No
6 weeks
Developing country
No
4 weeks
(Sub) tropical areas
No
Between 3 days and 1 month
All countries
Yes
1—3 weeks
Endemic area
No
Between 7 and 21 days
All countries
No
3 months
All countries
No
6 weeks
All countries
No
30 days
Canada [19]
France [10] Israel [20] Netherlands [21]
Netherlands [12]
New Zealand [16]
Singapore [23]
Switzerland [11]
USA [24] USA [13]
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
4
J. Freedman et al.
symptomatic cases reported through the enhanced surveillance scheme between 2007 and 2011 inclusive, where both dates were known. For the purpose of this study, the time between overseas travel and the onset date is defined from the date of return to or arrival in the UK and includes those whose trip originated from the UK and those who were new entrants or visitors to the UK. Cases were excluded from the analysis if they: had more than one organism isolated (at the species level only; different phage and antibiotic resistance types were not considered in this study), had no onset or travel dates, had onset before departure from the UK, or had onset before the minimum quoted incubation period after arrival in an endemic country (one day for paratyphoid and three days for typhoid). Cases reported between 2007 and 2010 were compared with those reported in 2011 in terms of the time between the onset date and overseas travel and the correspondence between the investigator classification and the operational travel-associated definition. The Mann—Whitney U test was calculated at the 5% level (twotailed) using IBM SPSS Statistics Version 19 to assess whether there were any differences in the non-normally distributed time intervals between the onset date and the dates of overseas travel between cases of typhoid and paratyphoid A during 2007—2010 and between 2007—2010 and 2011. The two-independent samples z test at the 95% level was calculated using STATA (STATA/SE 12.0 for Windows, Stata Corp LP., USA) to assess the significance of the difference between the proportions of cases classified as travel-associated in the two time periods. It is assumed that all of the investigators administered the questionnaires identically and that all of the cases had an equal chance of being interviewed and of answering all of the questions.
Table 2
Results Completeness of reporting Table 2 shows the completeness of reporting for the onset and travel history of cases during 2007—2010 and 2011. Reporting was much improved in 2011 with 83.2% of travel-associated cases, as designated by the investigators, having travel dates and onset date, compared to 65.2% during 2007—2010.
Cases reported between 2007 and 2010 Between 2007 and 2010, 1966 laboratory-confirmed symptomatic cases of typhoid and paratyphoid A and B (single isolates only) were reported in EWNI (an average of 492 each year). Of the cases with a complete travel history (N = 1845), 1725 (93.5%) were classified as travel-associated by the investigator, and 123 (6.5%) were classified as nontravel-associated. The non-travel-associated cases comprised the following: 37 specifically stated they had not traveled at all; 24 had traveled to an endemic country but the onset was before departure from the UK or occurred more than 21 days after overseas travel; and for the remaining 62 cases, it was unknown whether there had been no travel or if there had been travel outside of the incubation period. This information was not specifically requested on the questionnaire. Between 2007 and 2010, after the exclusion criteria were applied, a total of 1274 cases of travel-associated enteric fever (678 cases of typhoid, 549 cases of paratyphoid A and 47 cases of paratyphoid B) were included for analysis of time intervals between overseas travel and onset; 24% (297/1227) of typhoid and paratyphoid A cases became ill while they were abroad (Fig. 1). The majority of the remaining cases (460/511, 90.0%
Completeness of reporting travel and onset dates for enteric fever cases: 2007—2011.
Variables
Total symptomatic cases Travel history of symptomatic cases as designated by investigator Travel-associated cases designated by investigator Travel-associated cases designated by investigator: With travel dates With onset date With both travel and onset dates With both travel and onset dates and excluding cases with onset before travel
N cases (% total) 2007—2010
2011
1966 1845 (93.5)
481 477 (99.2)
1725 (87.4)
448 (93.1)
1262 (64.2) 1324 (67.3) 1281 (65.2) 1274 (64.8)
374 (77.8) 408 (84.8) 400 (83.2) 394 (81.9)
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever
5
Figure 1 Time (days) between onset date and arrival in the UK by organism with cumulative proportions for the cases designated as travel-associated (N = 1274, 65% of the total cases): 2007—2010.
typhoid; 359/419, 85.7% paratyphoid A) became ill within 21 days after overseas travel and a further 64 (25/511, 4.9% typhoid and 39/419, 9.3% paratyphoid A) became ill between 22 and 28 days after overseas travel, with a median of eight days for typhoid and nine days for paratyphoid A. There was no statistically significant difference in the distribution of time intervals between onset and overseas travel between typhoid and paratyphoid A (p = 0.12). For paratyphoid B, 43% (20/47) of cases became ill while they were abroad, and the remainder (27/47, 57%) became ill within 21 days after overseas travel, most (19/27, 70%) within the first seven days (median four days).
Cases reported in 2011 In 2011, 481 laboratory-confirmed symptomatic cases of typhoid and paratyphoid A and B (single isolates only) were reported in EWNI. Of cases with a completed travel history (477), 448 (93.9%) were designated as travel-associated by the investigator, and 29 (6.1%) were designated as nontravel-associated. The non-travel-associated cases comprised the following: nine cases specifically stated they had not traveled at all; one had traveled to an endemic country, though not recently (no dates were given); for the remaining 19 cases, it was unknown whether there had been no travel, or if there had been travel outside of the incubation period. When the exclusion criteria were applied, a total of 394 travel-associated cases of enteric fever
(203 cases of typhoid, 184 cases of paratyphoid A and seven cases of paratyphoid B) were included for analysis of time between onset and overseas travel; 21% (80/387) of the typhoid and paratyphoid A cases became ill while they were abroad, whereas 43% (3/7) of the paratyphoid B cases became ill while they were abroad (Fig. 2). The majority of the remaining cases (151/159 typhoid 95%; 141/148 paratyphoid A 95%; 3/7 paratyphoid B 43%) became ill within 28 days of overseas travel, with a median of seven days for typhoid and nine days for paratyphoid A. There was no statistically significant difference in the distribution of the time intervals between onset and overseas travel between typhoid and paratyphoid A (p = 0.41). There were no significant differences in the distribution of cases of typhoid and paratyphoid A in regards to the interval between onset and overseas travel in 2011 compared to 2007—2010 (p = 0.98 for typhoid and paratyphoid A); the distribution for paratyphoid B was similar in 2011, except for one case of paratyphoid B with onset of symptoms 56 days after return to the UK (Fig. 2).
Classification of travel-associated and non-travel-associated cases before and after the change in definition There was no significant difference in the proportion of total cases classified as travel-associated and non-travel-associated after the definition of a travel-associated case changed from those with
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
6
J. Freedman et al.
Figure 2 Time (days) between the onset date and arrival in the UK by organism with cumulative proportions for the cases designated as travel-associated (N = 394, 82% of total cases): 2011.
onset within 28 days compared to those within 21 days of overseas travel (Table 3). If the case definitions in use for each time period had been strictly applied to designate the travel status of the case, then the number of potentially misclassified UK acquired cases would have increased by 111 (90.2%) in 2007—2010 and by 16 (55.2%) in 2011. It is unclear from the information given on the surveillance forms whether further investigations were undertaken to seek a potential UK source in the cases that did not strictly fit the case definition; however, all but two of these cases were known to have recently arrived from or had traveled to a known endemic country, most on the Indian subcontinent. One case classified as travelassociated outside the case definition in 2008 had traveled to Germany, a country not typically known to be endemic for enteric fever as classified by NaTHNaC, and one case in 2007 had no country of travel stated.
Discussion There are a number of challenges associated with surveillance of travel-associated infections in developed countries [26], in particular the ascertainment of complete travel and clinical histories and the absence of an international ‘gold standard’ case-definition for travel- and non-travelassociated cases. Enhanced surveillance has greatly improved the completeness and quality of the data collection on enteric fever in these respects and has enabled a more accurate assessment of the enteric fever situation in EWNI [2]. Although enteric fever is primarily a travel-associated infection in the UK, sporadic cases associated with secondary transmission from a travel-associated case [28] and small clusters that are presumed to have been acquired in the UK have infrequently occurred [29—31]. More rigorous and time-consuming investigation and public health action is required for
Table 3 Designation of enteric fever as travel-associated or not by investigators, pre and post the change in definition of a travel-associated case: 2007—2011. Designation of cases by investigator
2007—2010
2011
% Change between 2007—2010 and 2011
Travel-associated cases (% of total)
1725 (93.5) (95% CI: 92.4—94.6) 123 (6.5) (95% CI: 5.4—7.6) 1845
448 (93.9) (95% CI: 91.8—96.0) 29 (6.1) (95% CI: 4.0—8.2) 477
Increased by 0.4% — not significant (p = 0.75)
Non travel-associated cases (% of total)
Total with known travel history
Decreased by 0.4% — not significant (p = 0.75)
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever non-travel-associated cases; therefore, it is important that cases be classified appropriately. One means of appropriate classification has been on the basis of a temporal relationship between symptom onset and travel. The initial definition of a travel-associated case as symptom onset within a three-week (21 day) period of overseas travel was not always adhered to by investigators. Differing incubation periods stated in various UK guidelines for investigating enteric fever in use at the time [25,32] might partly account for the differing assessments, but not strictly adhering to the initial definition might also reflect the fact that ‘time since travel’ is only one factor that investigators consider in assessing cases (for example, 4% of cases in 2007—2010 and 2011 were designated as travel-associated when they had onset after 28 days from overseas travel). Although there are a number of other questions included in the surveillance questionnaire [2] to ascertain whether a case might have been acquired in the UK, the reason for the designation as travel-associated or not is not always clear. This study shows that changing the definition of travel-associated cases of enteric fever in terms of the time between symptom onset and overseas travel did not significantly affect the proportion of cases designated by investigators as travel-associated. Our analysis suggests that time might be subordinate to other considerations when investigators classify a case, if, for example, there is no other plausible source of infection for the individual other than likely exposure abroad, especially in cases in which travel to an endemic country is known. This finding is consistent with a Norwegian study that considered the influence of time between return to Norway and symptom onset for other gastrointestinal (GI) infections; the study concluded that clinicians are likely to take into account the usual epidemiology of the infection within the home country when assessing whether a GI infection is travel-associated even if there is a long period between return to the home country and symptom onset [33]. A high proportion of typhoid and paratyphoid fever cases in EWNI have traveled abroad to visit friends and relatives [27], and many travel together in family groups; therefore, it is possible that secondary cases within a traveling group might be mistaken for primary travel-associated cases, which may account for the longer apparent incubation period in some cases. The purpose of a time frame is to provide guidance to investigators on which cases are very likely to be travel-associated and to help identify the cases in which further consideration should be given to the likelihood of them being UK-acquired (i.e., to prompt the investigation of other possible
7 risk factors), which is of significance because UK acquired cases require higher levels of public health action. One of the concerns that might be raised about extending the time frame for defining a case as travel-associated is that non-travel-associated cases might be consequently misclassified and not appropriately investigated and managed, leading to clusters of indigenous disease. There is no evidence from the enhanced surveillance that the change in definition has led to more UK-acquired cases of enteric fever being misclassified in 2011 than in previous years. The limitations of this study predominantly relate to the completeness of data capture, although this aspect has improved over time. In particular, for most cases classified as non-travelassociated, the basis on which this classification was made was not clear (e.g., because of the temporal relationship between travel and symptoms or because there had been no travel); this information is not requested on the form, but it might be worth revising the questionnaire to include this information for a future audit. Limited information was provided on the surveillance form about possible sources for non-travel-associated infections. Some sources are likely to be as a result of contact with known cases within a household; between 2007 and 2011, 14% (22/152) of non-travel-associated cases were likely to have been infected from a household contact who might or might not have traveled themselves. Subsequent to this study, and in part based on its findings, new ‘Public Health Operational Guidelines for Enteric Fever (Typhoid and Paratyphoid)’ [34] were published in early 2012. These guidelines aim to improve the investigation of nontravel-associated cases, providing algorithms for investigators to use in trying to identify potential sources of infection in the absence of a travel history within 28 days of the onset of symptoms. Part of the evaluation of these guidelines [35] includes a detailed analysis of cases known or suspected to be acquired in the UK to assess the value of more thorough investigation of cases with a long period of time between overseas travel and the onset date. It is anticipated that data completeness will be much improved subsequent to the implementation of these guidelines. For those deciding the public health actions required in response to a case of enteric fever, the temporal definition provides useful guidance (in conjunction with other information about the case) as to what public health actions are required. It might be that once the investigation of the case is complete and also considering the possibility of any contact with co-travelers, travel abroad is the most likely source in the absence of evidence to the
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
8
J. Freedman et al.
contrary, even if the case had onset of symptoms more than 28 days after arrival in or return to the UK. For surveillance reporting and data analysis purposes, it is appropriate to continue to rely on the professional judgment of the public health practitioner to define a case as travel-associated (or presumed travel-associated) as is currently the case [2,28]. In this analysis, the vast majority of enteric fever cases reported in EWNI between 2007 and 2011, in which the individuals did not become ill while they were abroad, had onset dates within 28 days of arrival in or return to the UK, with a median of nine days and seven days for paratyphoid A and typhoid, respectively, and a median of five days for paratyphoid B. The data demonstrate that, although the median for typhoid was slightly lower, there was no significant difference between typhoid and paratyphoid A in this respect, suggesting similar incubation periods for these organisms; however, the data suggest that paratyphoid B has a shorter typical incubation period. The upper limit of the typical incubation period quoted in the literature for enteric fevers is lower than that observed in this case series. This difference is especially striking because this analysis measures time from arrival in the UK to onset as a proxy for the incubation period, rather than the time from infection to onset, so that the actual maximum incubation periods observed in this study are likely to be longer. For enteric fever, as well as many other infections, incubation periods vary between individuals, depending on the infecting dose and the immune status and general health of the individual who has been infected. Using the duration between travel and symptom onset is only one factor to consider in determining whether or not a case is travel-associated.
Author contributions Joanne Freedman was responsible for the concept and design of the study, the analysis and interpretation of the data, and the drafting of and final approval of the article. Lorraine Lighton was responsible for advising on the interpretation of the data, critical appraisal of the article for the intellectual content and final approval of the article. Jane Jones was responsible for advising on the concept and design of the study and data interpretation, critical appraisal of the article for the intellectual content and final approval of the article.
Conflicts of interest Funding: No funding sources. Competing interests: We have no conflict of interest to declare. Ethical approval: Not required.
Acknowledgments The authors gratefully acknowledge the following: Our health protection colleagues in local PHE Centres and the Environmental Health Officers who completed the questionnaires for the enhanced enteric fever surveillance; our colleagues in the PHE Salmonella Reference Service for providing the data on laboratory-confirmed cases of enteric fever for the enhanced surveillance; Edgar Wellington, Travel and Migrant Health Section, who managed the daily administration of the enteric fever enhanced surveillance; Helen Green from the PHE Respiratory Diseases Department for providing statistical advice.
Conclusion Epidemiology is a complex science, and although evidence-based guidelines for best practice are useful in investigating cases of infectious disease, public health professionals must use their professional judgment when a case does not appear to be consistent with the rules. The key is to be able to gather as much information as possible about the case for accurate and consistent reporting and analysis of the data. Surveillance will continue to provide evidence to inform and improve the public health management of cases and to improve the advice given to travelers who might be at increased risk of acquiring enteric fever through overseas travel.
References [1] Maskey AP, Day JN, Tuan PQ, Thwaites GE, Campbell JI, Zimmerman M, et al. Salmonella enterica Serovar Paratyphi A and S. enterica Serovar Typhi cause indistinguishable clinical syndromes in Kathmandu, Nepal. CID 2006;42(May (1)):1247—53. [2] Public Health England (PHE). Enhanced surveillance of enteric fever webpage. Available on the HPA legacy website http://www.hpa.org.uk/Topics/InfectiousDiseases/ at: InfectionsAZ/TravelHealth/GeneralInformation/trav30 Enhancedsurveillanceofentericfever/ [accessed 10.06.13]. [3] Glynn JR, Bradley DJ. The relationship between infecting dose and severity of disease in reported outbreaks of Salmonella infections. Epidemiol Infect 1992;109:371—88. [4] Miner JR. The incubation period of typhoid fever. J Infect Dis 1922;31:296—301.
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
JIPH-296; No. of Pages 9
ARTICLE IN PRESS
Travel-associated cases of enteric fever
9
[5] Gillespie S. Salmonella Infections. In: Cook GC, Zumla A, editors. Manson’s tropical diseases. Edinburgh: Saunders; 2003. p. 937—49. [6] Mintz E, Sodha S. Typhoid fever. In: Heymann DL, editor. Control of communicable diseases manual. Washington: American Public Health Association; 2008. p. 664—71. [7] Crump JA, Luby SP, Mintz ED. The global burden of typhoid fever. Bull World Health Organ 2004;82:346—53. [8] World Health Organization. Diarrhoeal diseases: typhoid fever. Geneva World Health Organization Initiative for Vaccine Research; 2009. Available at http://www.who.int/ vaccine research/diseases/diarrhoeal/en/index7.html [last accessed 5.02.13]. [9] European Centre for Disease Prevention and Control (ECDC). Typhoid/paratyphoid fever. In: ECDC, editor. Annual epidemiological report 2011. Stockholm: ECDC; 2011. p. 116. [10] Delmas G, Vaillant V, Jourdan N, Le Hello S, Weill FX, de Valk H. Typhoid and paratyphoid fever in France between 2004 and 2009. Bull Epidemiol Hebd 2011;2:9—12. [11] Keller A, Frey M, Schmid H, Steffen R, Walker T, Schlagenhauf P. Imported typhoid fever in Switzerland, 1993 to 2004. J Travel Med 2008;15:248—51. [12] Bovee LP, van Kessel RP, Peerbooms PG, van den Hoek JA. Buiktyfus in Amsterdam, 1991—2000, en vereenvoudiging van het landelijk protocol voor bron- en contactopsporing [Typhoid fever in Amsterdam, 1991—2000, and the simplification of the national protocol for source and contact tracing]. Ned Tijdschr Geneeskd 2002;146: 1833—7. [13] Lynch MF, Blanton EM, Bulens S, Polynak C, Vojdani J, Stevenson J, et al. Typhoid fever in the United States, 1999—2006. JAMA 2009;302:859—65. [14] Taylor M, MacDougall L, Li M, Galanis E. The impact of international travel on the epidemiology of enteric infections, British Columbia, 2008. Can J Public Health 2010;101:332—6. [15] Skull SA, Tallis G. An evidence-based review of current guidelines for the public health control of typhoid in Australia: a case for simplification and resource savings. Aust N Z J Public Health 2001;25:539—42. [16] The Institute of Environmental Science and Research Ltd., Porirua, New Zealand Notifiable and other diseases in New Zealand: annual report 2010; 2011. [17] Shigematsu M, Okabe N. Trends of statutory reporting of travel associated typhoid fever, paratyphoid fever, cholera, and HAV infection in recent five years in Japan. In: Poster session presented at: 9th conference of the international society of travel medicine (CISTM9). 2005. [18] Ty AU, Ang GY, Ang LW, James L, Goh KT. Changing epidemiology of enteric fevers in Singapore. Ann Acad Med Singapore 2010;39:889—98. [19] Ravel A, Nesbitt A, Marshall B, Sittler N, Pollari F. Description and burden of travel-related cases caused by enteropathogens reported in a Canadian community. J Travel Med 2011;18:8—19.
[20] Meltzer E, Sadik C, Schwartz E. Enteric fever in Israeli travellers: a nationwide study. J Travel Med 2005;12:275—81. [21] Baaten GG, Sonder GJ, Van Der Loeff MF, Coutinho RA, van den HA. Fecal-orally transmitted diseases among travelers are decreasing due to better hygienic standards at travel destination. J Travel Med 2010;17:322—8. [22] Ministry of Health New Zealand. Communicable disease control manual 2012. Wellington: Ministry of Health; 2012. [23] Yew FS, Goh KT, Lim YS. Epidemiology of typhoid fever in Singapore. Epidemiol Infect 1993;110:63—70. [24] Ackers ML, Puhr ND, Tauxe RV, Mintz ED. Laboratory-based surveillance of Salmonella serotype Typhi infections in the United States: antimicrobial resistance on the rise. JAMA 2000;283:2668—73. [25] Hawker J, Begg N, Blair I, Reintjes R, Weinberg J. Communicable disease control handbook. 2nd ed. Oxford: Blackwell Science; 2005. [26] Lawrence J, Jones J, Hill DR. Improving the evidence base for pre-travel advice: the importance of surveillance of travel-associated infection. Br J Gen Pract 2005;55:566—8. [27] Health Protection Agency. Enteric fever epidemiological reports. Available at http://www.hpa.org.uk/web/ HPAweb&HPAwebStandard/HPAweb C/1259152344471 [28] Health Protection Agency. Enteric fever surveillance quarterly report (England, Wales and Northern Ireland): fourth quarter 2011. Health Protection Report 2012; 6: enteric. Available at http://www.hpa.org.uk/hpr/archives/2012/ hpr0612.pdf [29] Logan S, Jayasena C, Aali A, Ash S, Cooke FJ. Typhoid without travel. Clin Med 2010;10:299—300. [30] Public Health Laboratory Service. Typhoid fever outbreak in Newport, Wales: update. Commun Dis Rep CDR Wkly 2001;11(33), news. Available at http://www.hpa.org.uk/ cdr/archives/2001/cdr3301.pdf [31] Public Health Laboratory Service. Typhoid in North London. Commun Dis Rep CDR Wkly 1991;1(6), news. Available at http://www.hpa.org.uk/cdr/archives/1991/cdr0691.pdf [32] PHLS Advisory Committee on Gastrointestinal Pathogens. Preventing person-to-person spread following gastrointestinal infections: guidelines for public health physicians and environmental health officers. Commun Dis Public Health 2004;7:362—84. [33] Guzman-Herrador B, Vold L, Nygard K. Surveillance of travel-associated gastrointestinal infections in Norway, 2009—2010: are they all actually imported? Euro Surveill 2012;17(41), pii=20294. Available online at: http://www. eurosurveillance.org/ViewArticle.aspx?ArticleId=20294 [34] Health Protection Agency (HPA) and Chartered Institute of Environmental Health. Public health operational guidelines for enteric fever (typhoid and paratyphoid). London: HPA; 2012. [35] Balasegaram S, Potter AL, Grynszpan D, Barlow S, Behrens RH, Lighton L, et al. Guidelines for the public health management of typhoid and paratyphoid in England: practice guidelines from the National Typhoid and Paratyphoid Reference Group. J Infect 2012;65:197—213.
Available online at www.sciencedirect.com
ScienceDirect
Please cite this article in press as: Freedman J, et al. Defining travel-associated cases of enteric fever. J Infect Public Health (2014), http://dx.doi.org/10.1016/j.jiph.2013.07.011
