Journal of Internal Medicine 2006; 259: 410–417 doi:10.1111/j.1365-2796.2006.01629.x Rising incidence and persistently high mortality of hospitalized pneumonia: a 10-year population-based study in Denmark R. W. THOMSEN1, A. RIIS1, M. NØRGAARD1, J. JACOBSEN1, S. CHRISTENSEN1, C. J. MCDONALD2 & H. T. SØRENSEN1,3 From the 1Department of Clinical Epidemiology, Aarhus University Hospital, Aalborg and Aarhus, Denmark, 2The Regenstrief Institute and the Indiana University School of Medicine, Indianapolis, IN, USA, and 3Department of Epidemiology, School of Public Health, Boston University, Boston, MA, USA Abstract. Thomsen RW, Riis A, Nørgaard M, Jacobsen J, Christensen S, McDonald CJ, Sørensen HT. Rising incidence and persistently high mortality of hospitalized pneumonia: a 10-year populationbased study in Denmark. J Intern Med 2006; 259: 410–417. Background. Little is known about temporal trends in the incidence and mortality of pneumonia in the general population. Methods. We conducted a population-based cohort study in three Danish counties (population 1.4 million) to examine changes in the incidence and 30- and 90-day mortality associated with hospitalized pneumonia between 1994 and 2004. All adults hospitalized with a first-time diagnosis of pneumonia (n ¼ 41 793) were identified in hospital discharge registries and followed for mortality through the Danish Civil Registry System. We determined age-standardized incidence rates and adjusted mortality rates associated with calendar year, gender, age and comorbidity. Results. Between 1994 and 2003, the incidence of hospitalized pneumonia amongst adults increased Introduction Pneumonia is an important cause of morbidity and mortality and a heavy economic burden on healthcare systems, primarily due to hospitalized cases [1–3]. Some have suggested that the incidence of pneumonia may rise in Western populations due to the increased proportion of elderly persons with 410 from 288 per 100 000 person-years to 442 per 100 000 person-years, equivalent to an age-standardized incidence rate ratio of 1.50. The cumulative mortality within 30 and 90 days of admission was 15.2% and 21.9%, respectively, ranging from a 90-day mortality of 2.5% in patients aged 15–39 years to 34.7% in those aged 80 and over. Advanced age was the most important poor prognostic factor, followed by a high comorbidity score and male gender. The adjusted mortality rate ratios amongst patients with hospitalized pneumonia in 1999–2004, when compared with 1994–1998, were 0.89 (95% CI 0.85–0.94) after 30 days and 0.91 (95% CI 0.88–0.95) after 90 days. Conclusions. The incidence of hospitalized pneumonia in Denmark has increased considerably during the last 10 years and, combined with persistently high mortality rates, is of clinical and public health concern. Keywords: epidemiologic study, incidence, mortality, pneumonia. multiple morbidities [4, 5]. However, there is little population-based longitudinal data to evaluate temporal trends in the incidence and mortality of pneumonia [6, 7]. Reported in-hospital mortality rates of pneumonia average 5–15% and have apparently remained at this level for decades [8]. Mortality increases with age and comorbidity and probably male gender [9, 10], but population-based
2006 Blackwell Publishing Ltd HOSPITALIZED PNEUMONIA IN DENMARK studies on the effect of these prognostic factors are few [6]. We therefore conducted this populationbased study in three Danish counties to examine changes in the incidence and 30- and 90-day mortality associated with hospitalized pneumonia amongst adults between 1994 and 2004 and to assess the effect of age, gender and comorbidity on pneumonia survival. 411 pneumonia, we removed hospitalizations for any recurrent episode of pneumonia during 1994–2003 (n ¼ 19,874), and patients who were hospitalized for pneumonia during 1989–1993 (ICD-8 codes 480.XX–486.XX, 0.73.XX and 471.XX) (n ¼ 2,416). We finally removed paediatric patients <15 years of age (n ¼ 7,642). This gave us a population of unique adult patients with a first episode of pneumonia for analysis. Subjects and methods Validation of discharge diagnoses of pneumonia Study population and setting The study was conducted in the three Danish counties of North Jutland, Aarhus and Viborg, within a homogeneous Caucasian, mixed rural and urban population of 1.4 million inhabitants (26% of the Danish population). The Danish healthcare service provided tax-supported healthcare for all county residents, guaranteeing free access to hospitals and primary medical care. General practitioners (GPs)/family doctors acted as ‘gate keepers’ with regard to hospital treatment and treatment by specialists, and nearly all (97.6%) residents had registered with a specific GP of their choice throughout the study period [11]. Because the Danish government assigns a unique civil registry number to every Danish citizen, we could link patients from our population-based registries and establish a complete hospitalization history for each individual. Identification of hospitalized pneumonia We identified patients with a pneumonia associated hospitalization by searching the discharge diagnoses codes stored within the ‘Aarhus University Hospital Research Database’, which carries key information for all patients discharged from any hospital in the three counties since 1977 (in Viborg County since 1972). Diagnoses are coded by physicians according to the ‘International Classification of Diseases’, 8th revision (ICD-8) until the end of 1993 and 10th revision (ICD-10) thereafter. We first identified all discharges between 1 January 1994 and 31 December 2003 associated with a diagnosis of pneumonia (ICD-10 codes J12.x–J18.x), ornithosis (ICD-10 code A481.x), or legionellosis (ICD-10 code A709.x) (n ¼ 71,725). Because we were interested in the first episode of To evaluate the diagnoses of pneumonia identified in the database, we reviewed a random sample of 10 of the selected hospital records for each of the 10 years of the study period (1994 through 2003). We confined the review to North Jutland County, as data quality is considered uniform in the counties’ hospital discharge registries [12]. We confirmed the discharge diagnosis of pneumonia when the patient’s chest radiographic examination showed an infiltrate, and at least one of the following clinical or laboratory findings were present: body temperature ‡37.5 C, cough, dyspnoea, chest pain or rales coincident with the area of infiltrate, increased sputum, purulent sputum, microorganism isolated from blood culture, leucocyte count ‡12 · 109 L)1 or C-reactive protein >100 mg dL)1. We calculated the predictive value of a pneumonia discharge diagnosis as the percentage of episodes in the reviewed hospital record sample that fulfilled our criteria for confirmed pneumonia. We also assessed the proportion of community-acquired pneumonia episodes in the sample, i.e. episodes in which pneumonia was present or incubating at the time of hospital admission. Data on comorbidity To adjust for comorbidity, we computed the Charlson index score for each study subject on the basis of the complete hospital discharge history before the date of admission with pneumonia. The Charlson index includes 19 major disease categories, and has been adapted and validated for use with hospital discharge data in ICD databases for the prediction of short- and long-term mortality [13]. Three comorbidity index levels were defined, according to the Charlson index score: score of 0, low; 1–2, medium; 3+, high. As alcohol abuse is not included in the
2006 Blackwell Publishing Ltd Journal of Internal Medicine 259: 410–417 412 R . W . T H O M S E N et al. Charlson index and may influence pneumonia prognosis, we also collected data on previous alcohol-related disorders (‘yes’ or ‘no’) from the Discharge Registry (ICD-8 codes 291, 303, 979, 980, 577.10; ICD-10 codes F10, K86.0, Z72.1, R78.0, T51). Statistical analysis Incidence. The annual incidence rates of pneumonia were computed overall and within age- and genderspecific strata as the number of patients with a first discharge diagnosis of pneumonia divided by the number of citizens in the counties of North Jutland, Aarhus, and Viborg in the middle of the year (obtained from ‘Statistics Denmark’). Incidence rates were standardized to the population of North Jutland, Aarhus and Viborg County in 1994, using 1-year age groups. Mortality. For the mortality analysis, we linked patients with discharge diagnoses of pneumonia to the Danish Civil Registry System, which keeps records of all changes in vital status and migration and date of death for the entire Danish population. The follow-up time was computed from the date of hospital admission until death, migration, or 90 days after the admission date, which ever came first. We used Cox regression analyses to compute 30- and 90-day mortality rate ratios (MRRs) with 95% CIs comparing the first and second half of the study period (calendar year bands 1994–1998 and 1999–2003), gender (male/female), age group (15– 39, 40–64, 65–79, ‡80 years), and level of comorbidity (Charlson index score ¼ 0, 1–2, or 3+), whilst adjusting for all the other factors. We verified the assumption of proportional hazards in the Cox model graphically. All statistical analyses were performed with sas software (version 9.1.3, SAS Institute Inc., Cary, NC, USA). The study was approved by the Aarhus University Hospital Registry Board and the Danish Data Protection Agency (record no. 2004-41-3854). Results We identified 41 793 persons older than 15 years with a first time hospitalization with pneumonia. The median age at hospitalization was 73.6 years, and 53% of the individuals were male. Validation of discharge diagnoses of pneumonia In the reviewed sample of 100 episodes of hospitalized pneumonia, 90 episodes fulfilled criteria for confirmed pneumonia, equivalent to a positive predictive value of 90% (95% CI 82–95%). Of the 90 confirmed episodes, 78 (87%) were community acquired. Incidence Between 1994 and 2003, the incidence of hospitalized pneumonia increased from 288 per 100 000 person-years to 442 per 100 000 person-years, corresponding to an age-standardized incidence rate ratio of 1.50. The incidence increased similarly amongst young and old people over the decade, i.e. the incidence increased by 66% in the 15- to 39year-old age group and by 58% in the ‡80-year-old group. Overall, incidence rates were approximately 20% higher in men than in women throughout the study period, yet amongst persons aged 65 years or older, incidence rates in men were 50–100% higher than in women (Fig. 1). The incidence of hospitalized pneumonia rose more than 30-fold with age, i.e. from 84 per 100 000 in those aged 15–39 years to 2800 per 100 000 in those aged ‡80 years in 2003. Between 1994 and 2003, the proportion of hospitalizations in which pneumonia did not constitute the main diagnosis but rather a secondary discharge diagnosis increased from 32% to 36%. The median length of hospital stay for pneumonia patients decreased from 8 days (interquartile range, IQR 4–14) in 1994 to 7 days (IQR 3–12) in 2002. Mortality The overall cumulative mortality within 30 and 90 days of admission was 15.2% and 21.9%, respectively. Mortality was higher in patients who had pneumonia listed as a secondary discharge diagnosis (20.3% after 30 days) compared with a main diagnosis (12.6% after 30 days). As can bee seen in Fig. 2, mortality rose sharply with age – from 1.5% (30-day) and 2.5% (90-day) in patients aged 15–39 years to 24.7% (30-day) and 34.7% (90-day) in patients over age 80 years. This latter group (‡80 years) accounted for 47% of deaths after 90 days in our cohort (Table 1).
2006 Blackwell Publishing Ltd Journal of Internal Medicine 259: 410–417 HOSPITALIZED PNEUMONIA IN DENMARK 413 Fig. 1 Standardized incidence rates (SIR) of first-time hospitalized pneumonia in women and men according to age group. North Jutland, Aarhus, and Viborg County, Denmark, 1994–2003. Age standardized to the population of North Jutland, Aarhus, and Viborg County in 1994, using 1-year age groups. Fig. 2 Survival curves for patients with first-time hospitalized pneumonia according to age group. Mortality was also much higher in patients with a medium or high comorbidity score than in those with no registered comorbidities (Fig. 3). Table 1 shows adjusted 30- and 90-day MRRs according to calendar period, age group, gender and comorbidity score. A high level of comorbidity, and in particular advanced age, remained strong predictors of death. Adjusted mortality rates were 15–19% higher in men than in women. Inclusion of alcoholrelated disorders left the estimates virtually unchanged (data not shown). When we stratified analyses by calendar period, the MRRs for male gender, increased age and comorbidity were between 2% and 26% higher in 1999–2004 than in 1994–1998, but none of the differences reached statistical significance (data not shown). The overall mortality after hospitalized pneumonia was similar in 1999–2004 compared with 1994–1998. However, the proportion of pneumonia patients who were 80 years or older, and those with a high level of comorbidity, increased from 29% to 31% and from 15% to 18%, respectively. After
2006 Blackwell Publishing Ltd Journal of Internal Medicine 259: 410–417 MRR, mortality rate ratio. bAdjusted by Cox proportional hazards regression analyses for gender, age, comorbidity and calendar period. cLevel of Charlson index score, see text. a 1 (ref.) 1.45 (1.38–1.52) 2.07 (1.96–2.19) 1 (ref.) 1.79 (1.71–1.88) 2.62 (2.48–2.77) 1923 (10.5) 1497 (15.8) 2923 (21.0) 18 367 9487 13 939 1 (ref.) 1.69 (1.59–1.79) 2.34 (2.19–2.50) 1 (ref.) 1.37 (1.29–1.45) 1.86 (1.74–1.99) 2678 (14.6) 2104 (22.2) 4372 (31.4) 1 (ref.) 1.19 (1.14–1.24) 1 (ref.) 1.18 (1.13–1.22) 3935 (20.2) 5219 (23.4) 1 (ref.) 1.15 (1.10–1.21) 2765 (14.2) 3578 (16.0) 19 469 22 324 1 (ref.) 1.14 (1.08–1.20) (ref.) (3.33–5.21) (6.07–9.41) (10.74–16.63) 1 4.17 7.56 13.37 (ref.) (3.93–6.13) (7.93–12.25) (13.58–20.97) 1 4.91 9.86 16.87 (2.5) (11.6) (22.1) (34.7) 83 1114 3619 4338 (ref.) (3.59–6.41) (6.55–11.56) (11.83–20.86) 1 4.80 8.70 15.71 (ref.) (4.12–7.34) (8.21–14.45) (14.40–25.32) (1.5) (7.8) (15.0) (24.7) 49 749 2452 3093 3331 9576 16 386 12 500 1 5.49 10.89 19.09 1 (ref.) 0.96 (0.92–1.00) 4067 (22.4) 5087 (21.5) 1 (ref.) 0.89 (0.85–0.94) 1 (ref.) 0.93 (0.89–0.98) 2858 (15.7) 3485 (14.8) 18 187 23 606 Calendar year band 1994–1998 1999–2004 Age (years) 15–39 40–64 65–79 80+ Gender Female Male Comorbidity indexc Low (0) Medium (1–2) High (3+) Crude MRR (95% CI) Dead (%) Predictor n Dead (%) Crude MRRa (95% CI) Adjustedb MRR (95% CI) 90 day 30 day Table 1 Crude and adjusted 30- and 90-day mortality in hospitalized pneumonia according to calendar period, gender, age group and level of comorbidity 1 (ref.) 0.91 (0.88–0.95) R . W . T H O M S E N et al. Adjusted MRR (95% CI) 414 adjustment for these factors and minor gender differences, the relative mortality for hospitalized pneumonia in 1999–2004 compared with that in 1994–1998 decreased by 11% (95% CI 6–15%) after 30 days and 9% (95% CI 5–12%) after 90 days (Table 1). When we stratified on pneumonia as a main or secondary diagnosis, the 30-day mortality rate decreased by 15% (95% CI 9–21%) for secondary pneumonia and by 10% (95% CI 3–16%) for primary pneumonia. Discussion This large population-based study shows that the incidence of hospitalized pneumonia in a welldefined North European population has increased by 50% over the last decade. The increase occurred in all age groups and in both genders. During this same time period, 30- and 90-day mortality rates have improved but only slightly. The strengths of our study include the uniformly organized Danish public healthcare system that allowed a population-based design, with inclusion of all first time hospitalizations with pneumonia from all hospitals in the region and complete followup for mortality at 90 days. The Charlson index enabled us to adjust for most underlying diseases in the survival analyses, and the large study size provided good statistical precision of the estimates. Limitations to this study include the use of routine hospital discharge data to identify patients with pneumonia and comorbidities, as some coding errors may have occurred. Nevertheless, we found that the positive predictive value of hospitalized pneumonia was high compared with other discharge diagnoses. Misclassification of comorbidity may have lead to residual confounding and could affect our mortality estimates. The overall annual incidence of hospitalized pneumonia in our study of 2.9–4.4 per 1000 is remarkably similar to estimates from the few population-based studies in other industrialized countries, e.g. United States 1988–94 (3.7 per 1000) [1], Ohio 1991 (2.7 per 1000) [14], Germany 1998 (2.9 per 1000) [4], and Portugal 1998–2000 (2.7 per 1000) [15]. The age-specific incidence rates in 2003 for those aged 65–79 years (12.5 per 1,000) and those aged ‡80 years (28.0 per 1,000) also correspond well with the annual incidence rate of 18.3 per 1000 amongst Americans aged ‡65 years reported
2006 Blackwell Publishing Ltd Journal of Internal Medicine 259: 410–417 HOSPITALIZED PNEUMONIA IN DENMARK 415 Fig. 3 Survival curves for patients with first-time hospitalized pneumonia according to level of Charlson index score. by Kaplan et al. [6], and with Swedish incidence rates for hospitalized pneumonia of 14.4 per 1000 in persons aged 75–84 years and 26.3 per 1000 in those aged ‡85 years [16]. In contrast, our incidence rates were higher than those reported by Jackson et al. [7] (11.5 per 1000) in persons aged ‡65 years in Washington State, perhaps due to the fact that their study excluded nursing home residents. However, these studies were not designed to study temporal trends in pneumonia incidence. There are several plausible explanations for the rising incidence of hospitalized pneumonia. One might posit that physicians may have lowered their threshold for admitting pneumonia cases over the period of observation. If that was the case, we would expect a reduced mortality rate because a greater proportion of cases would be mild. However, we did not observe any substantial change in mortality over this time period. Alternatively, these results could come from more aggressive coding due to increasing economic incentives [17]. Again, the almost unchanged mortality of pneumonia over time argues against more meticulous case finding being the main explanation. It is unlikely that other changes in the Danish healthcare system accounted for the increase in pneumonia hospitalization rates. In Denmark as elsewhere, the vast majority of pneumonia patients will initially be examined, treated and followed outside the hospital setting. Patients with severe pneumonia symptoms, increased risk of complications due to age or comorbidities, or antibiotic treatment failure will be admitted to a public hospital in accordance with national guidelines [18], either by a GP or after direct contact with the hospital’s emergency room. There have been no other healthcare facilities than public hospitals for the treatment of pneumonia in our region during the study period, and there were only minimal changes in the number and distribution of available hospital beds in medical and surgical specialties in the period 1997–2003, when health statistics were electronically accessible from the Danish National Board of Health (available at http://www.sst.dk). The decreasing median length of hospital stay with pneumonia in our cohort followed a general trend for decreasing length of all registered hospitalizations in Denmark between 1994 and 2002 [19]. Demographic changes in our study population did not explain the rising incidence of hospitalized pneumonia. However, chronic diseases such as malignancies, diabetes, stroke sequelae, chronic obstructive pulmonary disease, and congestive heart failure have been associated with an increased pneumonia risk [5, 7, 20], and all of these conditions have become more prevalent in Denmark due to a number of factors including population ageing, lifestyle factors, obesity, and a longer disease survival. Other increasingly prevalent factors that might affect temporal trends in pneumonia prognosis include the number of surgical procedures performed, particularly in very elderly people, the use of immunosuppressive treatment including glucocorticoids, and the institutionalization of elderly people in environments that may promote the transmission of respiratory tract infections [21]. Only a few of these putative pneumonia risk factors have been established from population-based studies, and
2006 Blackwell Publishing Ltd Journal of Internal Medicine 259: 410–417 416 R . W . T H O M S E N et al. properly designed epidemiologic studies are needed to quantify the effect of these factors on pneumonia risk and outcome in general populations [7]. Several socio-economic and lifestyle-related factors have been associated with pneumonia, including smoking, crowding and obesity [22, 23]. Smoking appears to be one of the most important risks factor for pneumonia in both young and elderly adults [7, 22], but its prevalence has actually decreased in Denmark during the study period [24]. This may also hold true for crowding, whereas the prevalence of obesity has increased [25]. The results from a study by Baik et al. [22] suggested that women with a body mass index of 30 or greater have more than twice the risk of pneumonia than normal weight women. The higher incidence rate (and mortality) in men in our study is consistent with previous findings [7, 10, 14, 26], and the magnitude of the gender differences amongst elderly individuals in this study was remarkable. The specific mechanisms linking male gender with higher pneumonia risk and mortality have not been established but may include a higher prevalence of socio-economic and lifestyle-related risk factors. The combination of annual influenza vaccination and pneumococcal vaccination has been suggested to reduce the risk of subsequent hospitalization with pneumonia [27]. Exact data for influenza and pneumococcal vaccination rates in ‘at-risk’ groups in Denmark are unfortunately lacking, yet by contrast with many states of the US, annual influenza vaccination rates in our region probably have been well below 50% during the study period, whilst the overall administration of pneumococcal vaccine has been as low as 2 of 1000 people per year [28]. In the light of these differences in vaccination rates, the similarity of pneumonia incidence rates amongst elderly persons in our Danish region and the US is interesting. We found an overall 30-day mortality rate from hospitalized pneumonia of 15.2% which is higher than in some previous reports [9, 29] but in accordance with findings from other populationbased studies [2, 10, 15]. More than one-third of patients aged ‡80 years with hospitalized pneumonia in our region will not survive 3 months after admission. The sharply rising mortality rates after pneumonia with advanced age, and the stepwise increase in mortality with increasing levels of comorbidity corroborate previous findings [9, 10]. Nonetheless, it is of clinical concern that the 30-day mortality for those between 40 and 64 years in our study reaches about 8%, and this estimate parallels in-hospital mortality rates from pneumonia in American seniors between 65 and 69 years in a similar population-based study by Kaplan et al. [6]. Our findings suggest that hospitalized pneumonia is a very severe clinical condition even in middleaged adults with mortality rates paralleling those after acute myocardial infarction or stroke [30, 31]. The mortality after pneumonia improved slightly during the 10-year period of study. Changes in prognosis may be due to changes in diagnostic delay and treatment. Few aspects of pneumonia treatment are evidence based, but early and appropriate antibiotic therapy remains the mainstay in treatment [32]. Community-acquired pneumonia treatment guidelines have been virtually unchanged during the last decade in Denmark and include penicillin as first choice antimicrobial agent, as levels of penicillin resistance in the most frequently detected microbial agent, Streptococcus pneumoniae, remain uniquely low in Denmark [18]. In conclusion, the incidence of hospitalized pneumonia has increased considerably during the last decade in Denmark and, combined with persistently high mortality rates, is of great clinical and public health concern. Conflicts of interest statement No conflict of interest was declared. Acknowledgements This work was made possible through financial support from the ‘Klinisk Epidemiologisk Forskningsfond’, Denmark. References 1 Niederman MS, McCombs JS, Unger AN, Kumar A, Popovian R. The cost of treating community-acquired pneumonia. Clin Ther 1998; 20: 820–37. 2 Carriere KC, Jin Y, Marrie TJ, Predy G, Johnson DH. Outcomes and costs amongst seniors requiring hospitalization for community-acquired pneumonia in Alberta. J Am Geriatr Soc 2004; 52: 31–8. 3 Scott G, Scott H, Turley M, Baker M. Economic cost of community-acquired pneumonia in New Zealand adults. U933. N Z Med J 2004; 117: U933.
2006 Blackwell Publishing Ltd Journal of Internal Medicine 259: 410–417 HOSPITALIZED PNEUMONIA IN DENMARK 4 Welte T, Suttorp N, Marre R. CAPNETZ-community-acquired pneumonia competence network. Infection 2004; 32: 234–8. 5 Oosterheert JJ, Bonten MJ, Hak E, Lammers JW, Schneider MM, Hoepelman IM. The increase in pneumonia-related morbidity and mortality among adults in the Netherlands and possible explanations for it. Ned Tijdschr Geneeskd 2004; 148: 1765–9. 6 Kaplan V, Angus DC, Griffin MF, Clermont G, Scott WR, Linde-Zwirble WT. Hospitalized community-acquired pneumonia in the elderly: age- and sex-related patterns of care and outcome in the United States. Am J Respir Crit Care Med 2002; 165: 766–72. 7 Jackson ML, Neuzil KM, Thompson WW et al. The burden of community-acquired pneumonia in seniors: results of a population-based study. Clin Infect Dis 2004; 39: 1642–50. 8 Fine MJ, Smith MA, Carson CA et al. Prognosis and outcomes of patients with community-acquired pneumonia. A metaanalysis. JAMA 1996; 275: 134–41. 9 Mortensen EM, Kapoor WN, Chang CC, Fine MJ. Assessment of mortality after long-term follow-up of patients with community-acquired pneumonia. Clin Infect Dis 2003; 37: 1617–24. 10 Kaplan V, Clermont G, Griffin MF et al. Pneumonia: still the old man’s friend? Arch Intern Med 2003; 163: 317–23. 11 AnonymousHealth Care in Denmark. Copenhagen: Ministry of the Interior and Health, 2003. Available at: http:// www.im.dk/publikationer/healthcare_in_dk/index.htm 12 Nickelsen TN. Data validity and coverage in the Danish National Health Registry. A literature review. Ugeskr Laeger 2001; 164: 33–7. 13 de Groot V, Beckerman H, Lankhorst GJ, Bouter LM. How to measure comorbidity. A critical review of available methods. J Clin Epidemiol 2003; 56: 221–9. 14 Marston BJ, Plouffe JF, File TM et al. Incidence of communityacquired pneumonia requiring hospitalization. Results of a population-based active surveillance Study in Ohio. The Community-Based Pneumonia Incidence Study Group. Arch Intern Med 1997; 157: 1709–18. 15 Froes F. Pneumonia in the adult population in continental Portugal – incidence and mortality in hospitalized patients from 1998 to 2000. Rev Port Pneumol 2003; 9: 187–94. 16 Örtkvist Å. Pneumococcal disease in Sweden: experiences and current situation. Am J Med 1999; 107: 44S–9S. 17 Hsia DC, Krushat WM, Fagan AB, Tebbutt JA, Kusserow RP. Accuracy of diagnostic coding for Medicare patients under the prospective-payment system. N Engl J Med 1988; 318: 352–5. 18 Vestbo J, Benfield T. Pneumonibehandling. Rationel Farmakoterapi 2003; 12: 1–3. Available at: http://www.irf.dk. 19 Nye tal fra Sundhedsstyrelsen nr. 21, 2003. Landspatientregisteret gennem 25 år 1977–2002. Copenhagen: National Board of Health, 2003. 417 20 LaCroix AZ, Lipson S, Miles TP, White L. Prospective study of pneumonia hospitalizations and mortality of U.S. older people: the role of chronic conditions, health behaviours, and nutritional status. Public Health Rep 1989; 104: 350–60. 21 Marrie TJ. Community-acquired pneumonia in the elderly. Clin Infect Dis 2000; 31: 1066–78. 22 Baik I, Curhan GC, Rimm EB, Bendich A, Willett WC, Fawzi WW. A prospective study of age and lifestyle factors in relation to community-acquired pneumonia in US men and women. Arch Intern Med 2000; 160: 3082–8. 23 Nuorti JP, Butler JC, Farley MM et al. Cigarette smoking and invasive pneumococcal disease. Active Bacterial Core Surveillance Team. N Engl J Med 2000; 342: 681–9. 24 Anonymous. The Danish Health and Morbidity Survey 2000. Copenhagen: National Institute of Public Health, 2000. 25 Sørensen HT, Sabroe S, Gillman M et al. Continued increase in prevalence of obesity in Danish young men. Int J Obes Relat Metab Disord 1997; 21: 712–4. 26 Jokinen C, Heiskanen L, Juvonen H et al. Incidence of community-acquired pneumonia in the population of four municipalities in eastern Finland. Am J Epidemiol 1993; 137: 977–88. 27 Christenson B, Hedlund J, Lundbergh P, Örtqvist Å. Additive preventive effect of influenza and pneumococcal vaccines in elderly persons. Eur Respir J 2004; 23: 363–8. 28 Thomsen RW, Hundborg HH, Lervang HH, Johnsen SP, Schønheyder HC, Sørensen HT. Risk of community-acquired pneumococcal bacteremia in patients with diabetes: a population-based case-control study. Diabetes Care 2004; 27: 1143–7. 29 Torres OH, Munoz J, Ruiz D et al. Outcome predictors of pneumonia in elderly patients: importance of functional assessment. J Am Geriatr Soc 2004; 52: 1603–9. 30 Abildstrøm SZ, Rasmussen S, Rosen M, Madsen M. Trends in incidence and case fatality rates of acute myocardial infarction in Denmark and Sweden. Heart 2003; 89: 507–11. 31 Moon L, Moise P, Jacobzone S and the ARD-Stroke Experts Group. Stroke Care in OECD Countries: a Comparison of Treatment, Costs and Outcomes in 17 Countries. Paris: OECD, 2003. 32 Mandell LA, Bartlett JG, Dowell SF, File TM, Jr, Musher DM, Whitney C. Update of practice guidelines for the management of community-acquired pneumonia in immunocompetent adults. Clin Infect Dis 2003; 37: 1405–33. Correspondence: Reimar Wernich Thomsen MD, PhD, Department of Clinical Epidemiology, Aarhus University Hospital, Forskningens Hus, Sdr. Skovvej 15, Postbox 365, DK-9100 Aalborg, Denmark. (fax: +45 8942 4801; e-mail: uxreth@aas.nja.dk).
2006 Blackwell Publishing Ltd Journal of Internal Medicine 259: 410–417