Effect of Age on the Manifestations and Outcomes of Invasive Pneumococcal Disease in Adults

CLINICAL RESEARCH STUDY

Effect of Age on the Manifestations and Outcomes of Invasive Pneumococcal Disease in Adults Thomas J. Marrie, MD,a Gregory J. Tyrrell, PhD,b Sumit R. Majumdar, MD,c Dean T. Eurich, PhDd a

Department of Medicine, Dalhousie University, Halifax, Nova Scotia, Canada; bDivision of Diagnostic and Applied Microbiology, Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, Canada; cDepartment of Medicine, Faculty of Medicine and Dentistry, University of Alberta and The Provincial Laboratory for Public Health, Edmonton, Canada; dSchool of Public Health, University of Alberta, Edmonton, Canada.

ABSTRACT BACKGROUND: Although a considerable amount is known about the effect of age on the manifestations and outcomes of pneumonia, the same is not true for invasive pneumococcal disease. METHODS: This was a prospective observational study of all cases (2435) of invasive pneumococcal disease in adults in Northern Alberta from 2000 to 2014. Rates of invasive pneumococcal disease per 100,000, sociodemographic variables, clinical characteristics, and invasive pneumococcal disease–related outcomes were compared for the following age groups: 17-54, 55-64, 65-74, and ≥75 years. RESULTS: The rate of invasive pneumococcal disease per 100,000 increased with increasing age. Although only 27.3% of the cases were in those aged ≥65 years, they accounted for 48% of the deaths. The case fatality rate increased with increasing age, from 9.6% for those aged 17-54 years to 31.7% for those aged ≥75 years. The rate of meningitis decreased with increasing age, as did admission to intensive care and use of mechanical ventilation. There was a marked reduction in the rate of invasive pneumococcal disease due to protein conjugate vaccine 7 and protein conjugate vaccine 13 serotypes in those aged ≥55 years but a much smaller decline in rates for those aged 17-54 years. Replacement with non-vaccine serotypes constituted approximately 50% of the cases. CONCLUSIONS: The rate of invasive pneumococcal disease is highest in the very elderly, and manifestations of invasive pneumococcal disease are influenced by age. © 2018 Elsevier Inc. All rights reserved. • The American Journal of Medicine (2018) 131, 100.e1–100.e7 KEYWORDS: Age group; Invasive pneumococcal disease; Streptococcus pneumoniae

INTRODUCTION Funding: This study was supported by a grant-in-aid from Pfizer, Canada and Merck Canada Inc. The funders had no role in the design of the study or data analysis and they have not seen the manuscript. SRM holds the Endowed Chair in Patient Health Management supported by the Faculties of Medicine and Dentistry and Pharmacy and Pharmaceutical Sciences at the University of Alberta. DTE is a Canada Research Chair supported by the Government of Canada. Conflict of Interest: Other than the above 2 research grants—none. Authorship: TJM and GJT designed the study, organized the data collection, and had full access to all of the data in the study. DTE conducted all analyses and had full access to the data. All authors contributed to the interpretation of data, wrote the manuscript, revised the manuscript for intellectual content, and approved the manuscript to be published. Requests for reprints should be addressed to Thomas J. Marrie, MD, Dalhousie University, Department of Medicine, 59 Sprucebank Lane, RR 1 Chester Basin, Halifax, Nova Scotia B0J 1K0, Canada. E-mail address: [email protected]

0002-9343/$ - see front matter © 2018 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.amjmed.2017.06.039

There is a general understanding that the incidence of invasive pneumococcal disease is highest at the extremes of age.1 It is also well known that the mortality rate from invasive pneumococcal disease increases with increasing age.2 In pneumonia, it is well known that older patients report fewer symptoms than those who are younger.3 Whether this holds true for patients with invasive pneumococcal disease is uncertain because it is not well known how the incidence, predisposing factors, and outcomes (other than mortality) of invasive pneumococcal disease in adults vary by age group. There are 2 pneumococcal vaccines currently in use in Alberta, Canada. Protein conjugate vaccine 13 was implemented in the pediatric population for those individuals 2 months of age up to 59 months and for high-risk indivuduals 2 months of age up to and including 17 years of age in 2010.4

Marrie et al

Age Groups and Invasive Pneumococcal Disease

100.e2

Before 2010, protein conjugate vaccine 7 was in use and Identification and Serotyping of S. was introduced into the Alberta pediatric population in pneumoniae Isolates 2002 and discontinued when protein conjugate vaccine 10 was Streptococcus pneumoniae isolates were received at the Prointroduced. The second vaccine, pneumococcal polysacchavincial Laboratory from acute diagnostic laboratories in Alberta ride vaccine 23, has been in use since the late 1990s and is as per requirements of provincial notifiable disease regulaindicated for routine immunization in adults aged ≥65 years tions. These isolates were confirmed and those considered medically at as S. pneumoniae on the basis of risk.5 characteristic pneumococcal morCLINICAL SIGNIFICANCE We used data from a 15-year phology and optochin susceptibility population-based study of inva• In adults with invasive pneumococcal before serotyping.8 All pneumococsive pneumococcal disease in disease the mortality rate increases with cal isolates that exhibited a positive Northern Alberta, Canada to deQuellung reaction using commerincreasing age, from 9.6% to 31.7% for scribe the effect of age groups on cial type-specific antisera obtained those in the 17-54 years and ≥75 years the incidence, manifestations, outfrom Statens Serum Institute (Coage groups. comes, and prevalent serotypes in penhagen, Denmark) were assigned invasive pneumococcal disease. Fur• Symptoms of pneumonia as a manifesa serotype designation.9 Strains that thermore, we also explored the tation of invasive pneumococcal disease, were susceptible to optochin but that influence pneumococcal conjusuch as cough, sputum, and hemoptyfailed to serotype using the gated vaccines have had on the sis, decrease with increasing age. Quellung assay were assayed further prevalence of invasive pneumococusing the AccuProbe S. pneumoniae • Some of the processes of care, such as cal disease in specific age groups in culture identification test (Genadmission to intensive care and meour study population. Probe, San Diego, Calif) to confirm

METHODS Definitions

chanical ventilation, also decrease with increasing age.

• Changes in pneumococcal serotypes, especially in the very elderly, necessitate changes to current vaccination strategies.

Cases of invasive pneumococcal disease were defined as per the Canadian national case definition as isolation of Streptococcus pneumoniae from a sterile site, such as blood, cerebrospinal fluid, pleural fluid, biopsy tissue, joint aspiration, pericardial fluid, or peritoneal fluid. 6 Invasive pneumococcal disease is listed as a provincially notifiable disease in Alberta. Part of being a notifiable disease is the requirement that all invasive pneumococcal isolates be submitted to the Provincial Laboratory for Public Health for pneumococcal serotyping. This database allowed us to prospectively identify all cases of invasive pneumococcal disease in Northern Alberta.

Clinical Data Collection From 2000 to 2014, data were collected on all patients in Northern Alberta with invasive pneumococcal disease (approximate population 2,060,039). For each case, research nurses collected sociodemographic, clinical, functional, and laboratory data using a standardized case report form. The research nurses received training on data collection before the start of the study. Underlying illnesses were recorded according to the attending physician’s documentation in the medical record. Definitions of comorbidities and complications are as previously reported.7 This study received approval from the institutional research review committees of the Alberta Health Regions as well as the University of Alberta ethics review board.

the species identification.

Data Analysis

Sociodemographic variables, clinical characteristics, and invasive pneumococcal disease–related outcomes were compared according to age groups using Student’s t test or χ2 tests, or Fischer’s exact tests, as appropriate. For our age groups, we divided the adult population, those aged 17 years and older, into the following age groups: 17-54, 55-64, 65-74, and ≥75 years. In addition, for some analyzes we further subdivided the 17-54 year age group into 17-26, 27-36, 37-46, and 47-54 years of age. Incidence rates of invasive pneumococcal disease per 100,000 were calculated yearly from 2000 to 2014 across the age groups, as well as stratified wthin age groups by males and females. Specific comparisons between invasive pneumococcal disease rates across age and sex groups over time were evaluated using generalized linear models. Invasive pneumococcal disease serotypes were also identified and summarized according to age groups. We also further explored serotypes according to whether the invasive pneumococcal disease serotypes were included in the protein conjugate vaccine 7 and protein conjugate vaccine 13. All analyses were performed with Stata MP, version 14.2 (StataCorp, College Station, Tex).

RESULTS Overall, 2435 incident cases of invasive pneumococcal disease in adults occurred from 2000 to 2014. Table 1 shows the demographic features for the 4 age groups. There were major demographic changes from the youngest to the oldest groups. These included a significant decrease in the percentage of

100.e3

The American Journal of Medicine, Vol 131, No 1, January 2018

Table 1 Demographic Features and Selected Comorbidities of 2435 Adults with Invasive Pneumococcal Disease in Northern Alberta, According to Age Group Age Group (y) Variable

17-54

55-64

65-74

≥75

Overall, P Value

n (%) Age (y), mean (SD) Male Aboriginal Residence at home on admission Current smoker Illicit drugs Alcoholism Any comorbidity Alzheimer’s disease Stroke Congestive heart failure Coronary artery disease Atrial fibrillation Cirrhosis Hepatitis C Chronic renal failure HIV COPD

1304 (53.5) 40.6 (9.5) 784 (60.1) 244 (18.7) 1049 (80.4) 738 (56.6) 429 (33.1) 418 (32.1) 1069 (82) 2 (0.2) 17 (1.3) 12 (0.9) 38 (2.9) 5 (0.4) 98 (7.5) 258 (19.8) 28 (1.2) 111 (8.5) 84 (6.4)

464 (19.1) 59.6 (2.8) 263 (56.7) 45 (9.7) 394 (84.9) 221 (47.6) 52 (11.2) 137 (29.5) 414 (89.2) 3 (0.7) 15 (3.2) 14 (3) 56 (12.1) 13 (2.8) 41 (8.8) 46 (9.9) 26 (5.6) 6 (1.3) 106 (22.8)

289 (11.8) 69.9 (2.8) 151 (52.2) 15 (5.2) 261 (90.3) 92 (31.8) 1 (0.3) 44 (15.2) 271 (93.8) 11 (3.8) 25 (8.7) 34 (11.8) 80 (27.7) 32 (11.1) 8 (2.8) 2 (0.7) 23 (8.0) 0 102 (35.3)

378 (15.5) 82.5 (5.7) 182 (48.1) 8 (2.1) 256 (67.7) 53 (14.0) 0 21 (5.6) 367 (97.6) 67 (17.7) 49 (13) 87 (23) 158 (41.8) 78 (24.6) 4 (1.1) 1 (0.3) 50 (13.2) 0 149 (39.4)

<.001*†‡||¶ <.001†‡|| <.001*†‡§||¶ <.001*†‡§||¶ <.001*†‡§||¶ <.001*†§# <.001†‡§||¶ <.001*†‡§||¶ <.001†‡§||¶ <.001*†‡§|| <.001*†‡§|| <.001*†‡§||¶ <.001*†‡§||¶ <.001†‡§|| <.001*†‡§|| <.001*†‡||¶ <.001*# <.001*†‡§||

Values are number of cases (percentage) unless otherwise noted. COPD = chronic obstructive pulmonary disease; HIV = human immunodeficiency virus. *17-54 vs 55-64, P < .05. †17-54 vs 65-74, P < .05. ‡17-54 vs ≥75, P < .05. §55-64 vs 65-74, P < .05. ||55-64 vs ≥75, P < .05. ¶65-74 vs ≥75, P <.05. #Pairwise differences with ≥75 age group not estimatable owing to no events.

males and a decrease in percentage of aboriginal persons. In addition to the number of those residing at home, smoking, use of illict drugs, and alcoholism showed marked decreases from the youngest to the oldest age groups. As expected, notable differences in comorbidties were seen (Table 1) according to age group. With the exception of cirrhosis, hepatitis C, and human immunodeficiency virus infection, all increased with increasing age. The incidence of invasive pneumococcal disease was higher for individuals aged ≥75 years than for those aged 17-54 years. Figure 1 shows a comparison of the incidence of invasive pneumococcal disease per 100,000 males and females in the 17-54 and ≥75 year age groups. In the 17-54 year age group, the rate from 2000 to 2014 was essentially unchanged, with a higher rate observed in males than in females (P = .014). A peak in 2006 and 2007 is evident and has been previously reported as due to an outbreak of serotype 5 disease predominately affecting this age group.10 In the ≥75 year age group there is an overall decline in rates from 2000 to 2014 for both males and females. Unlike the 17-54 year age group, there is considerable fluctuation in rates of invasive pneumococcal disease year to year within the ≥75 year age group;

Figure 1 Incidence of invasive pneumococcal disease per 100,000 males and females in age groups 17-54 and ≥75 years. For 1754 year old males versus females, P = .014. For ≥75 year old males versus females, P = .033.

Marrie et al

Age Groups and Invasive Pneumococcal Disease

100.e4 individual serogroups showed a percentage increase from the youngest to the oldest; namely, 22F, 14, 11A, 6A, 6B, 33F, 23F, 16F, and 35B.

Figure 2 Incidence of invasive pneumococcal disease per 100,000 persons aged ≥55 years, according to pneumococcal serotypes contained in protein conjugate vaccines protein conjugate vaccine (PCV) 17 and protein conjugate vaccine 13 and serotypes not in these vaccines. Overall difference between the 3 categories, P = .0005. Protein conjugate vaccine 7 versus protein conjugate 13, P = .0007. Protein conjugate vaccine 7 versus other serotypes, P = .0004. Protein conjugate vaccine 13 versus other serotypes, P = .96.

however, the overall rate was similiarily higher in males than in females (P = .03). In 2014 the rate for males in the ≥75 year age group was 3.7 times higher than in the 17-54 year age group (30 vs 8 per 100,000), and for females it was 6.4 times higher (32 vs 5 per 100,000). Figure 2 illustrates the serotype changes for the ≥55 years of age population. Incidence rates drop significantly after 2002, when protein conjugate vaccine 7 was introduced for children, and in 2010, when protein conjugate vaccine 13 was introduced into this same population. It is important to note the increases seen for non-protein conjugate vaccine 13 serotypes up to 2014. The rate for protein conjugate vaccine 7 serotypes was lower than that for protein conjugate 13 (P = .0007) and for protein conjugate vaccine 7 verus other serotypes (P = .0004). There was no significant difference for rates of protein conjugate vaccine 13 serotypes versus other serotypes. The increase in non-protein conjugate vaccine 13 serotypes is driven primarily by a small number of serotypes. In 2014 alone, the last year reported, the incidence rates for the top 3 non-protein conjugate vaccine 13 serotypes for ages ≥17 years were 1.11 per 100,000 for serotype 20, 0.78 per 100,000 for serotype 22F, and 0.67 per 100,000 for serotype 8. Serotype 5 accounted for a high percentage of 17-54 age group disease owing to an outbreak of infection due to this serotype during the course of the study. Noteworthy is that overall only 21.7%-24.9% of isolates, depending on the age group, were in the protein conjugate vaccine 7 vaccine, and this increased to 40.7%-49.8% for protein conjugate 13. Nine

Outcomes. In this 15-year study 27.3% of the cases and 48% of the deaths from invasive pneumococcal disease in the adult population of Northern Alberta were in those aged ≥65 years. Table 2 gives the manifestations of invasive pneumococcal disease and some major treatments (intensive care unit admission and mechanical ventilation) according to age group. There was no change in the percentage of cases with pneumonia, whereas meningitis decreased significantly from the youngest to the oldest age group. The percentage in each group that had symptoms of cough and sputum decreased with increasing age. Surprisingly, although altered mental status was common in all groups, a significant increase was noted from the 17-54 to 55-64 year age groups, and after that it remained stable. The oldest age group had the lowest rate of admission to intensive care and use of mechanical ventilation compared with the other, younger groups. The complications and outcomes of invasive pneumococcal disease are given in Table 2, according to age group. The 30.7% rate of pleural effusion for the 17-54 year age group was significantly lower than that of the other age groups. Renal failure requiring dialysis was significantly lower in the oldest age group. Mortality showed a significant stepwise increase for each age group from youngest to oldest.

DISCUSSION It has been known for some time that the incidence of allcause pneumonia and invasive pneumococcal disease increases with advancing age.11,12 Although there has been a decline in the rate of invasive pneumococcal disease among those aged ≥75 years in the post-pneumococcal conjugate vaccine 13 era, the rate still remains higher than for those in the 17-54 year age group. This is supported by data from our surveillance study. In 2014 the rate for males in the ≥75 year age group was 3.7 times higher than that for males in the 17-54 year age group, whereas the corresponding figure for females was 6.4 times. We found no difference in overall incidence rates between 55-64 and 65-74 year olds or for males versus females in our survey. De St. Maurice et al12 also found no gender differences in rates for the 65-74 year old patients in the postpneumococcal conjugate vaccine 13 era. Many of the demographic changes that occur with age are not unexpected. Noteworthy is the high percentage of patients, 32.3%, not living at home. Most of these patients are in chronic care facilities, which may have implications for prevention of invasive pneumococcal disease. In one study 41% of patients with bacteremic pneumococcal pneumonia were from a nursing home.13 It has been previously documented that outbreaks of pneumococcal disease do occur in closed populations such as nursing homes.14 What is needed to better understand this entity are prospective studies of invasive pneumococcal disease in chronic care facilities.

100.e5

The American Journal of Medicine, Vol 131, No 1, January 2018

Table 2 Symptoms, Signs, Manifestations, and Complications of Invasive Pneumococcal Disease in 2435 Patients, According to Age Group Age Group (y) Variable

17-54

55-64

65-74

75 and over

Overall, P Value

Bacteremia Pneumonia Meningitis Cough Sputum Hemopytsis Altered mental status Mechanical ventilation Intensive care admission Pleural effusion Chest tube Empyema Acute aspiration Renal failure requiring dialysis Died

1246 (95.6) 1082 (82.3) 70 (5.4) 948 (72.7) 754 (57.8) 288 (22.1) 499 (38.3) 309 (23.7) 377 (28.9) 400 (30.7) 143 (10.9) 101 (7.7) 86 (6.6) 55 (4.2) 125 (9.6)

432 (93.1) 379 (81.7) 31 (6.7) 319 (68.8) 245 (52.8) 58 (12.5) 232 (50) 122 (26.3) 149 (32.1) 182 (39.2) 60 (12.9) 39 (8.4) 37 (8.0) 25 (5.4) 71 (15.3)

278 (96.2) 232 (80.3) 14 (4.8) 178 (61.6) 129 (44.6) 23 (8.0) 143 (49.5) 70 (24.2) 103 (35.6) 112 (38.8) 28 (9.6) 20 (6.9) 17 (5.9) 13 (4.5) 63 (21.8)

369 (97.6) 318 (84.1) 5 (1.3) 230 (60.8) 165 (43.7) 40 (10.6) 190 (50.3) 46 (12.2) 54 (14.3) 152 (40.2) 24 (6.3) 15 (4) 39 (10.3) 6 (1.6) 120 (31.7)

.015*|| .574 .005‡||¶ <.001†‡§|| <.001†‡§|| <.001*†‡ <.001*†‡ <.001‡||¶ <.001‡||¶ <.001*†‡ .016‡|| .056 .067 .041‡||¶ <.001*†‡§||¶

Values are number of cases (percentage). *17-54 vs 55-64, P < .05. †17-54 vs 65-74, P < .05. ‡17-54 vs 75 +, P < .05. §55-64 vs 65-74, P < .05. ||55-64 vs ≥75, P < .05. ¶65-74 vs ≥75, P < .05.

The increase in the various comorbidities that occurred with age was also not unexpected. What was unusual was that 82% of the 17-54 year age group had at least one comorbidity. In a previous study of invasive pneumococcal disease, 29% of adults in the 18-49 year age group and 46% of those in the 50-64 year age group had an indication for pneumococcal vaccination.15 Morbidities such as cirrhosis, hepatitis C, and human immunodeficiency virus infection decreased with increasing age in our population. This is likely a cohort effect because those aged ≥65 years 20 years ago were not in the prime age group for infection with hepatitis C and human immune deficiency virus infection. Twenty-four percent of the hepatitis C patients had cirrhosis, compared with 3.1% of those without hepatitis C.16 We found that the symptoms of cough and sputum production as well as hemoptysis decreased with increasing age. This is consistent with previous studies in pneumonia. For example, in a study of 1812 patients with all-cause communityacquired pneumonia categorized into 4 age groups—18-44, 45-64, 65-74, and ≥75 years—for 17 of the 18 symptoms reported there were significant decreases in reported prevalence with increasing age.3 For example, cough was reported by 90% of the 18-44 year age group and by 84% in the ≥75 year age group; dyspnea 75% and 61%; sputum production was reported by 64% of the patients in both age groups.3 Alteration of mental status was common at the time of presentation for all age groups in our cohort. The rate remained stable at approximately 50% for the age group ≥55 years. This is much higher than the 11.5% rate of altered mental status among adults with all-cause pneumonia admitted to Edmonton, Alberta hospitals.17 It is also higher than the 23% value

quoted in a review of sepsis associated encephalopathy.18 Because 30% of the oldest age group had Alzheimer’s or stroke we expected the rate to be highest in this group. All patients had a mini-mental status assessment, so we believe the results are reliable. Overall the mortality rate increased from 9.6% in the youngest age group to 31.7% in the oldest, which is not unexpected. There is a marked drop in the rate of admission to intensive care and use of mechanical ventilation for the oldest age group. This is in keeping with other studies of all-cause pneumonia and reflects the complexity of decision making in elderly patients with invasive pneumococcal disease.11 We observed a decrease in the rate of pneumococcal meningitis complicating invasive pneumococcal disease with increasing age. A decrease in pneumococcal meningitis among patients with invasive pneumococcal disease was seen in France after the introduction of protein conjugate vaccine 1319; however, this was not seen in the United States.20 It is also possible that, owing to increasing patient complexity with age, these patients were less likely to have cerebrospinal fluid collected, and hence some cases of pneumococcal meningitis may have been not diagnosed. An increase in empyema has been noted in the pediatric age group after the introduction of pneumococcal conjugate vaccines,21,22 but we did not see any significant changes in the rates of empyema across the adult age groups in this study. The rate of renal failure requiring dialysis was lowest in the oldest age group, but that was likely a treatment choice based on many other factors, such as advance directives. The mortality rate increased from 9.6% in the youngest age group to 31.7% in the oldest. We previously showed for

Marrie et al

Age Groups and Invasive Pneumococcal Disease

patients with invasive pneumococcal disease in Alberta, 10 factors were associated with increased 30-day mortality, and 2 were associated with lower 30-day mortality.7 The 10 were cancer within 5 years, cirrhosis, diabetes, requirement for oxygen supplementation, mechanical ventilation, altered mental status, cardiac arrest, infection with high or intermediated mortality serotypes, and treatment with a single antibiotic. The variables age 18-40 years and treatment with 2 antibiotics were found to be associated with lower mortality. We also examined factors predictive of mortality in a subset of the present study population.2 The following were associated with increased mortality: age per decade (adjusted odds ratio 1.42); nursing home residence, 3.34; dementia, 3.27; alcoholism, 2.32; multilobe pneumonia, 2.53; and guideline discordant antibiotic therapy, 3.24. Smoking (adjusted odds ratio 0.52) and infection with low case fatality rate serotypes (adjusted odds ratio 0.54) were associated with decreased mortality. The pneumococcal serotype dynamics in this study are most interesting. The effect of protein conjugate vaccine 7 and protein conjugate vaccine 13 was much greater in the ≥55 year age group compared with those aged 17-54 years. Indeed in the older age group the reduction in rates of infection due to protein conjugate vaccine 7 serotypes of 91.6% and 80% for the protein conjugate vaccine 13 serotypes is comparable to the rates achieved in young children.20 Part of this decline in the rates of invasive pneumococcal disease among the elderly has been attributed to the effectiveness of protein conjugate vaccine 7 and protein conjugate 13 vaccines in young children, thereby reducing the spread of pneumococci to older adults.23 The introduction of protein conjugate vaccine 7 and protein conjugate 13 vaccines led to rapid increases in serotype diversity among infecting strains of pneumococci, and then a decrease as a new equilibrium was established.24,25 This phenomenon of serotype replacement after the introduction of protein conjugate vaccine 7 led to an increase in serotypes 1, 3, 5, 6A, 6C, 7F, 12, 19A, and 22F.23 Six of the nine serotypes (1, 3, 5, 6A, 7F, and 19A) are in the protein conjugate vaccine 13 formulation. We also noted an increase in the percentage of 9 serotypes in the elderly groups, only 3 of which are in current protein conjugate vaccine formulations—serotypes 6A, 6B, and 23F. Indeed slightly less than half of the serotypes causing invasive pneumococcal disease in the very elderly are in current protein conjugate vaccine formulations. A new 15 valent protein conjugate vaccine containing strains 22F and 35 is currently undergoing clinical trials.26 It is likely, however, that new approaches to pneumococcal vaccines will be necessary to prevent the majority of cases of invasive pneumococcal disease as serotype changes continue to occur. Although many of our findings are what would be expected from what we already know of invasive pneumococcal in the elderly, the decline in the rate of meningitis and stability in the rate of impairment of mental status beyond age 55 years were unexpected. The emergence of replacement serotypes now causing 50% of invasive pneumococcal disease in the elderly will require a different approach to pneumococcal vaccination.

100.e6

ACKNOWLEDGMENTS We thank Ibrahim Quazi and Lilly Yusho for data management; Carol Mangan, RN, for data collection; and the microbiology staff of hospitals in Northern Alberta for forwarding Streptococcus pneumoniae isolates.

References 1. Backhaus E, Berg S, Anderson R, et al. Epidemiology of invasive pneumococcal infections: manifestations, incidence and case fatality rate correlated for age, gender and risk factors. BMC Infect Dis. 2016;16:367. 2. Beatty JA, Majumdar SR, Tyrrell GJ, Marrie TJ, Eurich DT. Prognostic factors associated with mortality and major in-hospital complications in patients with bacteremic pneumococcal pneumonia. Medicine (Baltimore). 2016;95:e5179. 3. Metlay JP, Schulz R, Li YH, et al. Influence of age and symptoms at presentation in patients with community-acquired pneumonia. Arch Intern Med. 1997;157:1453-1459. 4. Available at: http://www.albertahealthservices.ca/assets/info/hp/cdc/ifhp-cdc-pneu-13-conjate-vac-bio-pg-07-291.pdf. Accessed April 26, 2017. 5. Available at: http://www.albertahealthservices.ca/assets/healthinfo/hiflu-clinical-guidelines-pneumococcal-polysaccharide.pdf. Accessed April 26, 2017. 6. Canada Communicable Disease Report. Case definitions for diseases under national surveillance. Health Canada. 2000;26(S3):51. 7. Marrie TJ, Tyrrell GJ, Garg S, Venderkooi OG. Factors predicting mortality in invasive pneumococcal disease in Alberta. Medicine (Baltimore). 2011;90:171-179. 8. Facklam RR, Washington JA. Streptococcus and related catalasenegative gram-positive cocci. In: Balows A., Hausler W.J. Jr, Herrmann K.L., Isenberg H.D., Shadomy H.J., eds. Manual of Clinical Microbiology. 5th ed. Washington, DC: American Society for Microbiology; 1991:238-257. 9. Lund R, Henrichsen J. Laboratory diagnosis, serology and epidemiology of Streptococcus pneumoniae. In: Bergan T., Norris J.R., eds. Methods in Microbiology, Vol. 12. New York: Academic Press; 1978:241-262. 10. Tyrrell GJ, Lovgren M, Garg S, et al. A Streptococcus pneumoniae serotype 5 disease epidemic in western Canada: 2005-2009. Emerg Infect Dis. 2012;18(5):733-740. 11. Marrie TJ, Huang J. Epidemiology of community-acquired pneumonia in Edmonton, Alberta. An emergency room based study. Can Respir J. 2005;13:139-142. 12. de St Maurice A, Schaffner W, Giffin MR, et al. Persistent sex disparities in invasive pneumococcal disease in the conjugate vaccine era. J Infect Dis. 2016;214:792-796. 13. Murphy TF, Fine BC. Bacteremic pneumococcal pneumonia in the elderly. Am J Med Sci. 1984;288:114-118. 14. Tan CG, Ostrawski S, Bresnitz EA. A preventable outbreak of pneumococcal pneumonia among unvaccinated nursing home residents in New Jersey during 2001. Infect Control Hosp Epidemiol. 2003;24:848852. 15. Klemets P, Lyytikainen O, Ruutu P, Ollgren J, Nuroti JP. Invasive pneumococcal infections among persons with and without underlying medical conditions: implications for prevention. BMC Infect Dis. 2008;8:96. 16. Marrie TJ, Tyrrell GJ, Majumdar SR, Eurich DT. Prevalence, correlates, and outcomes of concurrent infection with hepatitis c and invasive pneumococcal disease. Emerg Infect Dis. 2017;In Press. 17. Huang JQ, Hooper PM, Marrie TJ. Factors associated with length of stay in hospital for suspected community acquired pneumonia. Can Respir J. 2006;13:317-324. 18. Zampieri FG, Park M, Machado FS, Azeuedo LCP. Sepsis associated encephalopathy and not just delirium. Clinics (Sao Paulo). 2011;66:18251831.

100.e7 19. Alari A, Chaussade H, De Celles M, et al. Impact of pneumococcal conjugate vaccine on pneumococcal meningitis cases in France between 2001 and 2014: a time series analysis. BMC Med. 2016;14(1):211. 20. Moore MR, Link-Gelles R, Schaffner W, et al. Effect of 13 valent conjugate vaccine in children on invasive pneumococcal disease in children and adults in the USA: analysis of multisite, population based surveillance. Lancet Infect Dis. 2015;15:301-309. 21. Kelly MM, Shadman KA, Edmonson MB. Treatment trends and outcomes in US hospital stays of children with empyema. Pediatr Infect Dis J. 2014;33(5):431-436. 22. Deceuninck G, Quach C, Panagopoulos M, et al. Pediatric pleural empyema in the province of Quebec: analysis of a 10-fold increase between 1990 and 2007. J Pediatric Infect Dis Soc. 2014;3(2):119-126.

The American Journal of Medicine, Vol 131, No 1, January 2018 23. Griffin MR, Zhu Y, Moore MR, Whitney CG, Grijalva CG. U.S. hospitalizations for pneumonia after a decade of pneumococcal vaccination. N Engl J Med. 2013;369:155-163. 24. Stockman C, Ampofo K, Pavia AT, et al. Clinical and epidemiological evidence of the Red Queen hypothesis in pneumococcal serotype dynamics. Clin Infect Dis. 2016;63:619-626. 25. Cilloniz C, Amaro R, Torres A. Pneumococcal vaccination. Curr Opin Infect Dis. 2016;29:187-196. 26. Golden AR, Adam HJ, Zhanel GC, Canadian Antimicrobial Resistance Alliance (CARA). Invasive Streptococcus pneumoniae in Canada, 2011-2014: characterization of new candidate 15-valent pneumococcal conjugate vaccine serotypes 22F and 33F. Vaccine. 2016;34:25272530.