Diagnostic value of procalcitonin for bacterial infection in elderly patients – a systemic review and meta-analysis S.-H. Lee,1,2 R.-C. Chan,1,2 J.-Y. Wu,3,4 H.-W. Chen,1 S.-S. Chang,4,5,6 C.-C. Lee7,8


Department of Rehabilitation and Physical Medicine, Taipei Veteran General Hospital, Taipei, Taiwan 2 Department of Medicine, College of Medicine, National Yang Ming University, Taipei, Taiwan 3 Department of Emergency Medicine, Chang Gung Memorial Hospital, Keelung, Taiwan 4 College of Medicine, Chang Gung University, Taoyuan, Taiwan 5 Department of Family Medicine, Chang Gung Memorial Hospital, Taoyuan, Taiwan 6 Graduate Institute of Clinical Medical Sciences, College of Medicine, Chang Gung University, Taoyuan, Taiwan 7 Department of Emergency Medicine, National Taiwan University Hospital, Yunlin Branch, Douliou, Taiwan 8 Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA Correspondence to: Shy-Shin Chang No. 5, Fusing St, Gueishan Township, Taoyuan County, 33305, Taiwan Tel.: +88633281200 (ext. 2505) Fax: +88633287715 Email: [email protected] Chien-Chang Lee, MD No. 95, 19th Neighborhood, Lianshih, Huwei Township, Yunlin County, 63247, Taiwan Tel.: + 886-23123456 (ext. 5926) Fax: +886-23223150 Email: [email protected] Disclosure None declared.



Review criteria

Objective: To summarise evidence for the diagnostic accuracy of procalcitonin (PCT) tests for identifying systemic bacterial infections in elderly patients. Methods: Major databases, including MEDLINE, EMBASE and the Cochrane Library, were searched for studies published from 1975 to March 2013 that evaluated PCT as a marker for diagnosing systemic bacterial infections in elderly patients and that provided sufficient data to construct two-by-two tables. Results: Four studies were available for quantitative meta-analysis. The area under a summary receiver operating characteristic curve was 0.89 (95% CI: 0.86–0.92). The overall sensitivity and specificity estimates for PCT tests were 0.83 (95% CI: 0.38–0.98) and 0.83 (95% CI: 0.60–0.94), respectively. These studies reported heterogeneous sensitivity estimates ranging from 0.24 to 0.96. The positive likelihood ratio for PCT (LR+ = 4.77; 95% CI: 2.49–9.13) was not sufficiently high for its use as a rule-in diagnostic tool, while its negative likelihood ratio was acceptably low for its use as a rule-out diagnostic tool (LR = 0.20; 95% CI: 0.04–0.97). Conclusions: Existing data suggest that PCT tests may add to the diagnosis of sepsis in elderly patients. We did not observe the performance of the PCT test in elderly patients inferior to adult patients. Given the imperfect accuracy, we do not recommend that the PCT test be used in isolation; instead, we suggest that it be interpreted in the context of clinical findings.

Introduction Infectious diseases are a common cause of increased morbidity and mortality in elderly patients and are often accompanied by serious complications, such as frequent recurrence, bloodstream infection and organ failure (1–6). Advanced age is associated with a decline in immune function commonly referred as ‘immune senescence’. Immune senescence is generally characterised by chronic, low-grade, systemic inflammation and impaired responses to immune challenge (1,3,4). Specifically, many vaccine- and pathogen-specific immune responses wane with age, while autoimmunity and inflammation increased in seniors. The production of procalcitonin (PCT) is induced by proinflammatory cytokines such as IL-1, IL-6 and TNF-alpha and expressed by nearly all kinds of parenchymal cells. It is unknown how the increased inflammation and declined functional reserve of parenchymal cells in elderly people would affect the response of PCT. Clinically, because of inadequately mounted immune response, the presentation of bacteraemia is

A predefined search strategy, using the PUBMED, EMBASE, and Cochrane databases, was employed to find studies that measured the accuracy of procalcitonin in the elderly. The study followed the PRISMA guidelines for systematic reviews, and the quality of each of the extracted studies was assessed using QUADAS criteria that had been developed for a diagnostic study meta-analysis.

Message for the clinic

• • •

The currently available data suggest that PCT tests may add to the diagnosis of sepsis in elderly patients. In a population with a low prevalence of sepsis, the PCT test can be used to help exclude sepsis in elderly people. We suggest that the PCT test be interpreted in the context of clinical findings.

different in elderly patients as compared with younger patients. Lack of fever or a mild fever, irregular leucocytosis and atypical clinical manifestations have been described with a higher prevalence as the subtle clinical manifestations of elderly bacteraemia (1,2,4– 6). Underdiagnosis of sepsis in the elderly population may lead to a delay in therapeutic intervention with a poorer outcome. PCT has been shown to have the best discriminative capability to differentiate bacterial infection from other non-bacterial infections or noninfectious inflammation (7–9). Randomised control trials have also showed that a PCT-guided algorithm can be used to safely reduce antibiotics use in patients with lower respiratory tract infection (10,11). These studies, however, have mostly been carried out in adult or paediatric populations, and none has examined the elderly population. The aim of this study was to systemically review and quantitatively evaluate the overall diagnostic accuracy of a PCT assay for bacterial infections in elderly patients. We also compared the diagnostic value of the PCT test in relation to the currently

ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357. doi: 10.1111/ijcp.12278

Procalcitonin in elderly patients

most widely used biomarker for detection of severer bacterial infection, C-reactive protein (CRP).

Methods Search strategy We conducted a systematic literature search for PCT and elderly people using Medline (via PubMed) and EMBASE from 1975 to November 2011 and updated the search until March 2013. PubMed was searched by combining two separate queries composed of the medical subject heading (MeSH) and text word (tw) keywords. PCT has not been listed as a MeSH term yet, so we used ‘procalcitonin’ as a text word without setting any language restrictions for the first query. The second query was performed using exploded headings and independent terms for elderly people: ‘aged’ [MeSH term] OR aged [tw] OR elderly [tw]. A similar search strategy and similar search terms were used in EMBASE, employing the EMbase TREE tool (EMTREE). The searches and studies included were not limited by publication date, country or language. The PubMed and EMBASE searches were conducted independently by two authors. To ensure comprehensive acquisition of literature, independent supplemental manual searches were performed on the reference lists of relevant articles and Cochrane databases.

Inclusion and exclusion Two reviewers independently identified articles eligible for in-depth examination using the following inclusion and exclusion criteria. Studies were included if at least one of the following outcomes was analysed: bacteraemia, sepsis, invasive bacterial disease and microbiologically or clinically documented bacterial infection. The relevant diagnostic tests were PCT assay alone or PCT assay compared with assessment of other laboratory markers such as CRP. The study types included were cross-sectional and case–control diagnostic test study designs, whereas case reports, review articles, editorials and clinical guidelines were excluded. We included studies on human subjects aged 65 years or older. We excluded mixed adult and elderly populations. Any discrepancies on articles meriting inclusion between reviewers were resolved by a consensus meeting of three authors. The study selection is summarised in Figure 1.

Data extraction and synthesis Data were extracted regarding the overall study characteristics (including the first author, country, language and year of publication); patient characteristics (including age range and clinical setting, patient population and disease group); numbers required for reconstruction of a 2 9 2 table (including number ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357

of participants, sensitivity, specificity and case number); and characteristics regarding the PCT test (cutoff levels), study setting and outcome definition (microbiologically documented infection, MDI or clinically documented infection, CDI). When studies were identified as containing pertinent data not included in the published article, we contacted the authors to obtain the missing data. When a response was not provided, such articles were excluded. The quality of the selected studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) criteria (12).

Data analysis We calculated the pooled sensitivity and specificity with 95% confidence intervals and plotted forest plots based on the reconstructed 2 9 2 contingency table of each individual study. We used the bivariate model for diagnostic meta-analysis to obtain weighted summary estimates of the sensitivity and specificity (13). The bivariate approach assumes a bivariate distribution for the logit-transformed sensitivity and specificity. By modelling paired sensitivity and specificity simultaneously, the bivariate model adjusts for the inherent negative correlation between the sensitivity and specificity that arises from the different thresholds used in different studies. The bivariate model also adopts a random-effects approach for both sensitivity and specificity, which accounts for clinical and methodological heterogeneity between studies beyond chance. The overall sensitivity, specificity and their 95% confidence intervals were calculated based on the binominal distributions of the true positives and true negatives. A summary receiver operating characteristic curve was also constructed as a way to summarise the true- and false-positive rates from different diagnostic studies. To formally quantify the extent of between-study variations, we calculated the inconsistency index I2, which represents the proportion of heterogeneity not explained by random variation (14). Statistically significant heterogeneity was considered present at I2 > 50%. To explore the potential source of heterogeneity, we performed subgroup analysis on predefined variables, which included PCT cut-off value, study setting and outcome definition. We also performed a conventional diagnostic odds ratio meta-analysis. Unadjusted data were used exclusively in all meta-analyses. The presence and the effect of publication bias were examined using Egger tests (15,16). Statistical analyses were conducted using STATA 11.0 (Stata Corp, College Station, TX), notably with the midas and metandi commands. All statistical tests were twosided, and statistical significance was defined as a pvalue less than 0.05.



Procalcitonin in elderly patients

Figure 1 Flow chart of study identification and inclusion

Results Our search yielded 7833 citations. After screening the title and abstract with a focus on the age range of the study population, only seven studies were eligible for further full-text review, of which five met our inclusion criteria (Figure 1). Study by Dwolatzky et al. included healthy older home residents as the control group. We thought this study should not be included for quantitative meta-analysis as including healthy controls may over-evaluate the performance of the diagnostic test. Finally, the four included studies comprised 760 patients [median, 195 (range 108–262)]. Bacterial infection was diagnosed in 134 patients [overall prevalence, 15.83% (range 9–23%)] (Table 1). PCT tests were used in all four studies (17–20), and CRP and leucocyte count tests were evaluated in three of them (19–21). Figure 2 provides an overall picture of the methodological quality of the studies as evaluated by the

QUADAS tool. QUADAS is an instrument, which contains 14 items specifically developed to assess the quality of systematic reviews of diagnostic test studies. The quality of studies was generally high, with all studies meeting 10 or more of the criteria. All studies collected data prospectively, used the same reference standard for outcome verification independent of the index test and provided clear descriptions of the patient selection criteria and index test. All patients underwent either a blood culture test or a complete clinical evaluation as the reference test without differential disease ascertainment. However, none of the included studies provided an explanation of the uninterpretable test results or participants’ withdrawal from the study. Incorporation bias is likely, as none of the studies provided a description as to whether the outcome verification was done by blinding to the index test. Table 1 lists the characteristics of all four included studies. Three studies defined elderly people as those ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357


ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357

92.0 36.0 39.0 96.0 NA ED Microbiologically documented infection

Inpatient Microbiologically documented infection 0.23 (172)

0.18 (262)

≥ 75

≥ 75

Steichen, 2009, France (19)

Lai, 2010, Taiwan (18)

PCT, procalcitonin; CRP, C-reactive protein; ED, emergency department; NA, not available.

0.15 (218) ≥ 75 Stucker, 2005, Switzerland (20)


0.11 (108) ≥ 65 Caterino, 2004, USA (17)


Prevalence (n)



0.5 30 0.51 175 0.38

Clinically and microbiologically documented infection


93.0 57.0 24.0 94.0 64.0 94.0 96.0 63.2 ED Clinically and microbiologically documented infection PCT > 0.2

Outcomes Cut-off value (PCT, ng/ml CRP, mg/l) Biomarkers tested Author, year, country

We performed subgroup analysis (Table 2) by restricting studies with a similar cut-off value, study settings and age range. The sensitivity decreased substantially with correspondent rise in specificity. Other studies lowered the threshold value for PCT (0.2–0.38 ng/ml), and the changes in sensitivity and specificity reflect this fact. The two studies that only reviewed emergency department (ED) patients showed a high-pooled sensitivity (sensitivity: 0.97, 95% CI: 0.89–0.99) but a low-pooled specificity (specificity: 0.61, 95% CI: 0.56–0.67) (17,18). In contrast, the remaining two studies performed in the inpatient settings showed a low-pooled sensitivity (sensitivity: 0.46, 95% CI: 0.34–0.58) but a high-pooled specificity (specificity: 0.94, 95% CI: 0.90–0.96) (19,20).

Age range (years)

Subgroup analysis

Table 1 Summary of the characteristics of the included studies

The results of the meta-analysis showed that the PCT test is an equally sensitive and specific test in diagnosing severe bacterial infection among elderly patients, whereas the CRP test is a highly sensitive but non-specific test. Leucocyte count was a poorly sensitive and highly specific test. We plotted the summary receiver operating characteristics (ROCs) (Figure 3). The PCT test had an area under the ROC (AUROC) of 0.89 (95% CI: 0.86–0.92). The pooled sensitivity and specificity estimates for the PCT test were 0.83 (95% CI: 0.38–0.98) and 0.83 (95% CI: 0.60–0.94), respectively (Figure 4). The pooled sensitivity and specificity estimates for the CRP test were 0.91 (95% CI: 0.39–0.94) and 0.36 (95% CI: 0.31–0.96), respectively. The pooled sensitivity and specificity estimates for the leucocyte count were 0.26 (95% CI: 0.19–0.35) and 0.86 (95% CI: 0.82–0.89), respectively. Figure 4 shows the forest plot of sensitivity and specificity for the PCT test. The CIs for specificity were substantially smaller than those for sensitivity because more patients finally turned out to be not infected. There were considerable withinand between-study variations in both sensitivity and specificity. The positive likelihood ratio for the PCT test was 4.77 (95% CI: 2.49–9.13), and the negative likelihood ratio was 0.20 (95% CI: 0.04–0.97). A substantial degree of heterogeneity was observed for the PCT test (I2, 68.0%; 95% CI: 67.95–6.86) but not for the CRP test (I2, 0.0%; 95% CI: 0.0–89.6) or leucocyte count (I2, 0.0%; 95% CI: 0.0–89.6).


Results of individual studies


PCT Sensitivity (%) Specificity (%)

CRP Sensitivity (%), Specificity (%)

aged greater than 75 years, while the other one study used 65 years of age as the elderly group definition. The outcome definitions can be classified into two broad categories: MDIs and CDIs. The sensitivity and specificity of the PCT or CRP tests are also presented in Table 1.


Procalcitonin in elderly patients


Procalcitonin in elderly patients

Figure 2 QUADAS (Quality Assessment of Diagnostic Accuracy Studies) criteria for the included studies. The figure showed an overview picture of the quality of the included studies assessed by the 14 items of QUADAS instrument

Figure 3 Plot of sensitivity and specificity for studies using PCT for the detection of bacterial infection among elderly patients, together with the summary receiver operating characteristic (SROC) curve (solid line) and the bivariate summary estimate (solid square), and the corresponding 95% confidence ellipse (inner dashed line) and the 95% prediction ellipse (outer dotted line). The symbol size for each study is proportional to the study size

Discussion The geriatric population (≥ 65 years) represents the fastest growing segment of the general population. The

clinical presentations of sepsis in elderly people (65– 84 years of age) are quite different from those in the adult population (6). Atypical presentations of sepsis in the elderly population, such as consciousness disturbance, afebrile episodes, shock and blunt leucocytosis response, make early diagnosis a great challenge for clinicians (1,2,5). It has been suggested that immune senescence is responsible for the atypical clinical presentation (17,22). Whether inflammatory biomarkers remain a reliable diagnostic tool for systemic infection in the elderly population has not been studied in depth. In this regard, a meta-analysis may increase the statistical power and the precision of estimates. In our analysis, which included 760 geriatric patients with suspected bacterial sepsis, we showed that the PCT test has a suboptimal rule-in value for confirming the diagnosis of bacterial sepsis. The overall positive likelihood ratio (LR+: 4.77, 95% CI: 2.49–9.13), although not sufficiently high, but, can be a useful diagnostic test compared with the low LR+ of CRP or WBC. For example, in ED setting with a 17% prevalence (pretest probability) of bacterial sepsis, a positive likelihood ratio of 4.77 translates into a positive predictive value (posttest probability) of only 48% (17,18). In other words, approximately one in two patients with a positive PCT test result can be expected to have either clinically or microbiologically confirmed bacterial sepsis. In contrast, the diagnostic value for PCT to rule out bacterial sepsis in elderly patients is acceptably high. In the same population with a 17% prevalence of bacterial sepsis, a negative likelihood ratio of 0.20 translates into a 96% negative predictive value. In the ward setting where the prevalence of sepsis is ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357



2.48 (1.48–4.15)

Figure 4 Forest plot of the sensitivity (left) and specificity (right) for studies using PCT to detect bacterial infection among elderly patients

ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357

PCT, procalcitonin; CRP, C-reactive protein.

0.26 (0.19–0.35) 3

0.86 (0.82–0.89)

2.05 (1.40–3.01)

0.86 (0.74–1.00)

0.70 (0.55–0.85)

0.0 (0.0–89.6)

0.706 9.52 (4.45–20.4) 0.91 (0.39–0.94) 3

0.36 (0.31–0.96)

1.43 (1.28–1.60)

0.60 (0.47–0.76)

0.80 (0.61–0.91)

0.0 (0.0–89.6)

0.572 0.480 68.0 (67.95–6.86) 78.7 (31.6–93.4) 18.11 (5.32–61.6) 7.90 (2.16–29.0) 0.89 (0.86–0.92) 0.78 (0.56–0.91) 0.83 (0.38–0.98) 0.48 (0.37–0.59)

Procalcitonin Overall (17–20) Cut-off = 0.5 (17,19,20) CRP Overall analysis (19–21) WBC Overall analysis (17,19,20)

4 3

0.83 (0.60–0.94) 0.89 (0.85–0.92)

4.77 (2.49–9.13) 4.33 (1.51–12.38)

0.20 (0.04–0.97) 0.58 (0.32–1.05)

Diagnostic OR (95% CI) Likelihood ratio+ Specificity (95% CI) Sensitivity (95% CI) Variables

Number of studies

Table 2 Summary of subgroup analysis of the included studies by different study characteristics

Likelihood ratio

AUROC (95% CI)

I2 (95% CI)

Publication bias (Egger’s test p)

Procalcitonin in elderly patients

lower (15%), the post-test probability for a positive or negative PCT test result was 44% and 3%, respectively. The low posttest probability for a negative result makes PCT a suitable tool for ruling out of sepsis in the elderly patients. However, given the huge social and medical cost of a missed diagnosis of sepsis in geriatric patients, one may still challenge the usefulness of the PCT test as a rule-out screening tool. We admit that the PCT test cannot be used as a standalone test, but this does not mean that the PCT test adds little to the current diagnostic work up of suspected sepsis. Several clinical trials have shown that an algorithm that integrates clinical and biomarker information in interpreting intermediate results of the PCT test may be used to safely reduce the prescription of antibiotics in patients with lower respiratory tract infections (11,23,24), or can be used to safely shorten the duration of antibiotics treatment duration. Compared with a previous systematic review of the diagnostic performance of the PCT test as a marker of bacteraemia in normal adult populations, we do not find evidence of immune senescence for the PCT test in elderly patients. A previous meta-analysis showed a similar AUROC (0.84, 95% CI: 0.75–0.90), sensitivity (0.76, 95% CI: 0.66–0.84) and specificity (0.70, 95% CI: 0.60–0.79) to our report (8). For CRP, an earlier systemic review evaluating the accuracy of PCT and CRP tests for the diagnosis of bacterial infections in a normal adult population showed a pooled sensitivity of 0.75 (95% CI: 0.62–0.84) and a pooled specificity of 0.67 (95% CI: 0.56–0.77) (8). We found a similar trend of a high sensitivity and


Procalcitonin in elderly patients

low specificity of the CRP test in diagnosing sepsis in elderly patients. The superior sensitivity (0.91, 95% CI: 0.39–0.94) and inferior specificity (0.36, 95% CI: 0.31–0.96) as compared with the previous report cannot be confirmed with confidence because of the relatively small number of studies and the wide confidence intervals. Our study has both strengths and limitations. In addition to the conventional sensitivity, specificity and AUROC, we reported a LR in our study that allows the clinician to better interpret the results of the diagnostic test. Unlike a positive or negative predicative value, LR is not influenced by the prevalence of the disease in the tested population and can be compared across different studies. To increase the statistical power and precision of the meta-analysis results, we considered studies using clinical-based outcome definitions, which may introduce betweenstudy heterogeneity. We observed a high degree of heterogeneity in this meta-analysis, which may be partly explained by the wide variation in cut-off values in different studies. For example, Steichen et al. used a cut-off value as high as 0.51 ng/ml, which thus yielded higher specificities and lower sensitivities (19), whreas Caterino et al. used a low cut-off value (0.2 ng/ml) (17), which yielded opposite results. Although we set age over 65 years as the inclusion criteria, in fact, most of the included studies are over 75 years of age. About 50% of patients in the four included studies are over 75. Therefore, results of this study are more suitable to be generalised to

References 1 Ackermann RJ, Monroe PW. Bacteremic urinary tract infection in older people. J Am Geriatr Soc 1996; 44: 927–33. 2 Fernandez-Sabe N, Carratala J, Roson B et al. Community-acquired pneumonia in very elderly patients: causative organisms, clinical characteristics, and outcomes. Medicine (Baltimore) 2003; 82: 159–69. 3 Gutierrez F, Masia M. Improving outcomes of elderly patients with community-acquired pneumonia. Drugs Aging 2008; 25: 585–610. 4 Lee CC, Chang IJ, Lai YC, Chen SY, Chen SC. Epidemiology and prognostic determinants of patients with bacteremic cholecystitis or cholangitis. Am J Gastroenterol 2007; 102: 563–9. 5 Lee CC, Chen SY, Chang IJ, Chen SC, Wu SC. Comparison of clinical manifestations and outcome of community-acquired bloodstream infections among the oldest old, elderly, and adult patients. Medicine (Baltimore) 2007; 86: 138– 44. 6 van Duin D. Diagnostic challenges and opportunities in older adults with infectious diseases. Clin Infect Dis 2012; 54: 973–8.

patients aged 75 and older. Lastly, the current analysis does not address the added diagnostic value to current practice based on clinical information and routine laboratory tests. This question will need to be answered through well-designed prospective studies comparing the two diagnostic strategies. In conclusion, the currently available data suggest that PCT tests may add to the diagnosis of sepsis in elderly patients. We did not observe an inferior performance of the PCT test in elderly patients. In a population with a low prevalence of sepsis, the PCT test can be used to help exclude sepsis in elderly people. We do not recommend that the PCT test be used in isolation; instead, we suggest that it be interpreted in the context of clinical findings. Further clinical trials are needed to assess whether incorporation of the PCT test in practice may ultimately improve the outcome. Before then, these data should help to inform evidence-based practice guidelines on the management of sepsis in elderly patients.

Acknowledgement The authors thank Pei-Shan Hsieh from Medical Wisdom Inc. for assistance with the statistical analysis of data.

Funding This study was supported by National Science Council NSC 100-2341-B-002-138-MY3NSC grant.

7 Jones AE, Fiechtl JF, Brown MD, Ballew JJ, Kline JA. Procalcitonin test in the diagnosis of bacteremia: a meta-analysis. Ann Emerg Med 2007; 50: 34–41. 8 Simon L, Gauvin F, Amre DK, Saint-Louis P, Lacroix J. Serum procalcitonin and C-reactive protein levels as markers of bacterial infection: a systematic review and meta-analysis. Clin Infect Dis 2004; 39: 206–17. 9 Uzzan B, Cohen R, Nicolas P, Cucherat M, Perret GY. Procalcitonin as a diagnostic test for sepsis in critically ill adults and after surgery or trauma: a systematic review and meta-analysis. Crit Care Med 2006; 34: 1996–2003. 10 Long W, Deng X, Zhang Y, Lu G, Xie J, Tang J. Procalcitonin guidance for reduction of antibiotic use in low-risk outpatients with community-acquired pneumonia. Respirology 2011; 16: 819–24. 11 Schuetz P, Christ-Crain M, Thomann R et al. Effect of procalcitonin-based guidelines vs standard guidelines on antibiotic use in lower respiratory tract infections: the ProHOSP randomized controlled trial. JAMA 2009; 302: 1059–66. 12 Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25.

13 Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58: 982–90. 14 Lijmer JG, Bossuyt PM, Heisterkamp SH. Exploring sources of heterogeneity in systematic reviews of diagnostic tests. Stat Med 2002; 21: 1525–37. 15 L’Abbe KA, Detsky AS, O’Rourke K. Meta-analysis in clinical research. Ann Intern Med 1987; 107: 224–33. 16 Song F, Khan KS, Dinnes J, Sutton AJ. Asymmetric funnel plots and publication bias in meta-analyses of diagnostic accuracy. Int J Epidemiol 2002; 31: 88–95. 17 Caterino JM, Scheatzle MD, Forbes ML, D’Antonio JA. Bacteremic elder emergency department patients: procalcitonin and white count. Acad Emerg Med 2004; 11: 393–6. 18 Lai CC, Chen SY, Wang CY et al. Diagnostic value of procalcitonin for bacterial infection in elderly patients in the emergency department. J Am Geriatr Soc 2010; 58: 518–22. 19 Steichen O, Bouvard E, Grateau G, Bailleul S, Capeau J, Lefevre G. Diagnostic value of procalcitonin in acutely hospitalized elderly patients. Eur J Clin Microbiol Infect Dis 2009; 28: 1471–6.

ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357

Procalcitonin in elderly patients

20 Stucker F, Herrmann F, Graf JD, Michel JP, Krause KH, Gavazzi G. Procalcitonin and infection in elderly patients. J Am Geriatr Soc 2005; 53: 1392–5. 21 Dwolatzky T, Olshtain-Pops K, Yinnon AM et al. Procalcitonin in the elderly: normal plasma concentrations and response to bacterial infections. Eur J Clin Microbiol Infect Dis 2005; 24: 763–5.

ª 2013 John Wiley & Sons Ltd Int J Clin Pract, December 2013, 67, 12, 1350–1357

22 Pawelec G, Larbi A, Derhovanessian E. Senescence of the human immune system. J Comp Pathol 2010; 142(Suppl. 1): S39–44. 23 Bouadma L, Luyt CE, Tubach F et al. Use of procalcitonin to reduce patients’ exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet 2010; 375: 463–74.


24 Burkhardt O, Ewig S, Haagen U et al. Procalcitonin guidance and reduction of antibiotic use in acute respiratory tract infection. Eur Respir J 2010; 36: 601–7.

Paper received November 2012, accepted August 2013

Diagnostic value of procalcitonin for bacterial infection in elderly patients - a systemic review and meta-analysis.

To summarise evidence for the diagnostic accuracy of procalcitonin (PCT) tests for identifying systemic bacterial infections in elderly patients...
195KB Sizes 0 Downloads 0 Views