Infection is a common complication in acute stroke. Whether or not preventive antibiotics reduce the risk of infection or even lead to a favorable outcome and reduction of mortality after a stroke still remains equivocal. This review was performed to update the current knowledge on the effect and possible benefits of prophylactic antibiotic therapy in patients with stroke.
A systematic review and meta-analysis of preventive antibiotics`effect on the incidence of infection, favorable outcome (mRS≤2) and mortality in patients with acute stroke is performed with relevant randomized controlled trials.
Six studies were identified, involving 4125 participants. Compared with the control group, the treated groups were significantly less prone to suffer from early overall infections [RR = 0.52, 95%CI (0.39, 0.70), p<0.0001], early pneumonia [RR = 0.64, 95%CI (0.42, 0.96), p = 0.03] and early urinary tract infections [RR = 0.35, 95%CI (0.25, 0.48), p<0.00001]. However, there was no significant difference in overall mortality [RR = 1.07, 95%CI (0.90, 1.27), p = 0.44], early mortality [RR = 0.99, 95%CI (0.78, 1.26), p = 0.92], late mortality [RR = 1.12, 95%CI (0.94, 1.35), p = 0.21] or favorable outcome [RR = 1.00, 95%CI (0.92, 1.08), p = 0.98].
Although preventive antibiotic treatment did reduce the occurrence of early overall infections, early pneumonia and early urinary tract infection in patients with acute stroke, this advantage was not eventually translated to a favorable outcome and reduction in mortality. Future studies are warranted to identify any subgroup of stroke patients who might benefit from preventive antibiotic treatment.
Citation: Zheng F, Spreckelsen Nv, Zhang X, Stavrinou P, Timmer M, Dohmen C, et al. (2017) Should preventive antibiotics be used in patients with acute stroke? A systematic review and meta-analysis of randomized controlled trials. PLoS ONE 12(10): e0186607. https://doi.org/10.1371/journal.pone.0186607
Editor: Giuseppe Vittorio De Socio, Azienda Ospedaliera Universitaria di Perugia, ITALY
Received: June 22, 2017; Accepted: October 4, 2017; Published: October 19, 2017
Copyright: © 2017 Zheng et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by the National Natural Science Foundation of China (81660211), http://www.nsfc.gov.cn/. Shengtao Yao received the funding, and works on Funding acquisition, Investigation, Methodology, Project administration and Validation of current research.
Competing interests: The authors have declared that no competing interests exist.
Infection is a common complication in the acute phase after a stroke and diagnosed in 21% to 65% of patients, with the most common infections being pneumonia and urinary tract infections [1–7]. Many studies have shown infections in patients with a stroke to be associated with poor short- and long-term functional outcome as well as a higher mortality [4, 8–10]. While this view is well established, a few studies failed to support this hypothesis [11, 12].
Several randomized controlled trials (RCTs) have assessed the preventive use of antibiotics in patients with acute stroke with conflicting results [11, 13–15]. A meta-analysis with pooled data from 506 stroke patients subsequently showed a 14% reduction in all infections with preventive antibiotics, but their effectiveness in decreasing post-stroke pneumonia, disability, and mortality was equivocal . Whether or not preventive antibiotics reduce the risk of poor functional outcome after stroke still remains uncertain . Due to the lack of sufficient evidence to prove it effective, existing guidelines do not lend support to the use of prophylactic antibiotics in patients who have had a stroke .
Recently, a systematic review  following the publications of two large RCTs updates the knowledge regarding the effect of prophylactic antibiotics treatment on post-stroke infections and the occurrence of mortality among adult acute stroke patients. Nevertheless, the effect of preventive antibiotics on early infections, early and late mortality, outcome in ischemic and hemorrhagic stroke, as well as the length of hospital stay still remain uncertain. [18–20]. Therefore, we conduct a systematic review and meta-analysis with randomized controlled trials to better understand the effect of preventive antibiotics in patients with acute stroke.
This systematic review and meta-analysis was conducted in accordance with PRISMA guidelines . All full text RCT, analyzing the incidence of infections, mortality, and functional outcome of preventive antibiotic therapy versus control (placebo or open control) in published studies were included. Case reports, non-randomized studies, comments, letters, editorials, protocols, guidelines, and animal studies were excluded. The Pubmed, Embase, and Cochrane library were searched for English-language articles published prior to Feb, 27, 2017. Unpublished studies were excluded. The data bases were searched using appropriate keywords and MeSH terms (See S3 File). Titles, abstracts, and subject headings were searched. The reference lists of all included articles and review papers were scrutinized for additional publications.
The primary outcome was any early infection, early pneumonia, early urinary tract infection (UTI), secondary outcome included late infection, overall mortality, early mortality, late mortality, the length of hospital stay, favorable outcome defined by the modified Rankin Scale (mRS) at the end of follow-up in patients with acute stroke, acute ischemic stroke and acute hemorrhagic stroke. Early infection was defined as occurrence of infection within the first 14 days after onset of symptoms. Early mortality was defined as occurrence of mortality within the first 14 days after onset of symptoms. Late infection was defined as occurrence of infection between early and the end of follow-up after onset of symptoms. Late mortality was defined as occurrence of mortality after 14 days or more after onset of symptoms. Overall mortality was defined as early mortality plus late mortality. Favorable outcome was defined as the mRS ≤2 at the end of follow-up.
Data collection and extraction
Two reviewers (F.Z. and X.Z.) independently screened the titles of all papers identified in the search to eliminate duplicate references, reviewed abstracts and selected RCTs according to the inclusion criteria.
The following information was extracted from the included studies: (1) general information: author names and publication year; (2) baseline information: study design, inclusion and exclusion criteria, patients’ characteristics, intervention characteristics, the definition of infection; (3) the data of infection: the number of early and late infection events at the end of follow-up, as well as the number of pulmonary- and urinary tract infections in patients with early infection; (4) the data of mortality: the number of early and late mortality events at the end of follow up (5) the data of functional outcome: the number of favorable functional outcomes at the end of follow-up. Due to different definitions of infection throughout the studies (See S1 File), we adopted the definition used by each investigator. If both data for intention-to-treat (ITT) and per protocol (PP) were available, the ITT data was used to avoid preselection bias and avoid data influenced by “drop outs” through early death in the PP-population. When algorithm-based diagnosis and physician-based diagnosis of post-stroke infections were both available, the algorithm-based diagnosis was used to minimize detection bias.
Analysis of the quality of the studies
The quality of the eligible studies was assessed using the Cochrane Collaboration’s tool for assessing risk of bias. Sequence generation, allocation sequence concealment, blinding, incomplete outcome data, selective outcome reporting and other potential sources of bias were evaluated. In accordance with the above mentioned tool, the risk for bias was set as either low, high or unclear (indicating either lack of information or uncertainty over the potential for bias) .
The data included in the present study is dichotomous, the relative risk (RR) and 95% confidence intervals (CI) were therefore calculated as the pooled mean effect size estimate using the Mantel–Haenszel method. The present meta-analysis was carried out using REVMAN software (version 5.3 for Windows. Copenhagen: The Nordic Cochrane Centre, the Cochrane Collaboration). Quality assessment of the included studies was done independently by two reviewers. Statistical heterogeneity was measured using c2 and I2 statistics. Due to the existence of heterogeneity in study design (double-blind versus open-label studies), type of antibiotic therapy (adequately covering all pathogens in post-stroke infections versus mostly chosen for neuroprotective properties) and definitions of infection, we chose a random-effects model for the pooled analyses. At the recommendation of the Cochrane Statistical Methods Group , a significance level of P value of heterogeneity was set at 0.1, and the I2 statistic was interpreted as follows: 0% to 40% as low heterogeneity; 30% to 60% as moderate heterogeneity; 50% to 90% as substantial heterogeneity; and 75% to 100% as considerable heterogeneity. Statistically significant heterogeneity was considered present at p<0.10 and I2>50%. In this case, sensitivity analysis, which excludes the study contributing most to the heterogeneity, was employed to identify the robustness of the result.
Following the literature search scheme and characteristics of the eligible references
As shown in Fig 1, we searched the Cochrane Library, PubMed database and EMBASE database and obtained an initial pool of 1677 references. Preliminary screening eliminated 221 duplicate references, and perusal of the titles and abstracts eliminated 1423 more references that did not meet the eligibility criteria. Examination of the full text eliminated 27 more references. These screening steps resulted in the identification of 6 studies involving a total of 4125 patients that were eligible for meta-analysis and systematic review [11, 13, 15, 19, 20, 24]. All patients included in these 6 studies were at least 18 years old. The characteristics of the 6 studies and the detailed outcome measurement are shown in Tables 1 and 2.
Analysis of the quality of eligible references
We evaluated the eligibility criteria of the 6 identified studies using the Cochrane Collaboration tool. The quality analysis of all 6 studies is shown in Fig 2. Briefly, randomization methods were described in all 6 studies [11, 13, 15, 19, 20, 24] and allocation concealments were adequate in 4 studies [11, 13, 15, 19]. For blinding, 6 studies used blind observers to assess outcome [11, 13, 15, 19, 20, 24], while blinded for carers or patients were used in 2 studies [11, 13]. In addition, 5 studies reported the complete outcome [13, 15, 19, 20, 24]. Finally, none study had selective outcome reporting.
In the six studies analyzing the risk to suffer from any early infection proved to be significantly lower in the antibiotics group [RR = 0.60, 95% CI (0.41, 0.88), p = 0.008]. Due to the existence of significant heterogeneity (p = 0.007; I2 = 69%), a sensitivity analysis was employed to assess the robustness of the findings. After exclusion of one trial by Kalra et al , the overall significance still remains [RR = 0.52, 95%CI (0.39, 0.70), p < 0.0001], without substantial heterogeneity (P = 0.28, I2 = 21%).
In the five studies which analyzed early pneumonia [13, 15, 19, 20, 24] initially no significant difference between both groups was found. [RR = 0.81, 95%CI (0.51, 1.31), p = 0.40]. However, in this subgroup the heterogeneity proved to be high (p = 0.08, I2 = 51%). After excluding the study performed by Kalra et al. (2015) which only included patients with dysphagia and contributed most to the heterogeneity, the heterogeneity lowered to (p = 0.86; I2 = 0%) and a significantly better outcome in the antibiotic group regarding early pneumonia was shown. [RR = 0.64, 95% CI (0.42, 0.96), p = 0.03].
Early urinary tract infection.
Five studies recorded the event of early urinary tract infection (UTI) [13, 15, 19, 20, 24]. A significant advantage for the group treated with preventive antibiotics was shown. [RR = 0.35, 95% CI (0.25, 0.48), p < 0.00001] without substantial heterogeneity (p = 0.46, I2 = 0%).
None of the defined secondary outcomes were positively influenced by the preventive use of antibiotics. The meta-analysis showed no significant difference in overall mortality, early mortality, late mortality, the length of hospital stay or favorable outcome. (Table 3, forest plots in online supplements).
The overall mortality at the end of follow-up was reported in all six studies [11, 13, 15, 19, 20, 24]. The number of overall mortality at the end of follow-up was 341 out of 2066 (17%) in the preventive antibiotics group versus 318 out of 2059 (15%) in the control group, with no significant difference [RR = 1.07, 95% CI (0.90, 1.27), p = 0.44], and no substantial heterogeneity. (p = 0.34; I2 = 12%).
Five studies recorded the event of early mortality in patients with acute stoke [13, 15, 19, 20, 24]. Both in the antibiotic group and the control group the early mortality rate was 6% (124 of 1999 vs. 125 of 1990) and no significant difference or heterogeneity was found. [RR = 0.99, 95% CI (0.78, 1.26), p = 0.92] (p = 0.84, I2 = 0%).
All six studies recorded the event of late mortality [11, 13, 15, 19, 20, 24]. In the preventive antibiotics group the number of late mortality events was 217 out of 2066 (11%) versus 193 out of 2059 (9%) in the control group. No significant difference regarding late mortality or substantial heterogeneity was found. [RR = 1.12, 95% CI (0.94, 1.35), p = 0.21] (p = 0.61, I2 = 0%).
The length of hospital stay.
Two studies analyzed the correlation between the length of hospital stay and the administration of prophylactic antibiotic treatment, which present the comparison of days in hospital between antibiotic group and control group as 26 (12–55) vs 19 (9–43) and 6 (3–10) vs 6 (3–11) [19, 20]. No significant difference in the length of hospital stay was found between the groups in either of the two studies, although further quantitative analysis was not possible in the present study due to unavailable detailed data.
Favorable outcome was defined as mRS ≤ 2 at the end of follow-up and was assessed in four studies [15, 19, 20, 24]. In both groups 47% of the patients had a favorable outcome. (919 of 1960 antibiotics vs. 917 of 1950 control) and no significant difference in outcome was found. [RR = 1.00, 95% CI (0.92, 1.08), p = 0.98], without substantial heterogeneity (p = 0.33, I2 = 10%).
Difference in outcome in hemorrhagic vs. ischemic stroke.
Three studies [15, 20, 24] analyzed the outcome for ischemic stroke and could not show a significant difference in favorable outcome (61%, 690 / 1127 antibiotic group vs. 59%, 674 / 1135 control group). [RR = 1.03, 95% CI (0.96, 1.10), p = 0.38], without substantial heterogeneity (p = 0.51, I2 = 0%).
Two studies [20, 24] reported hemorrhagic strokes and failed to show significant effect of antibiotics on favorable outcome as well. (63 out of 148 (41%) antibiotic group versus 61 out of 132 (46%), [RR = 0.92, 95% CI (0.71, 1.19), p = 0.52], without substantial heterogeneity (p = 0.73, I2 = 0%).
Infections are the most common medical complication occurring in patients with stroke and are considered to be a key cause of mortality . Pneumonia occurs in 11%-22%, and the risk of suffering from a UTI is even higher (16–24%) [6, 26]. The etiology of post-stroke pneumonia appears to be multifactorial. Invasive maneuvers, such as bronchoscopic diagnostics, aspiration of nasopharyngeal secretions caused by a decreased level of consciousness and reduced bulbar reflexes with oropharyngeal dysphagia [27–29], or feeding tube placement , could all contribute to a higher risk of pneumonia. In addition, old age, urinary incontinence, bedridden state with crural weakness, history of chronic obstructive pulmonary disease, diabetes, and stroke severity have been identified to be independent predictors of post-stroke nosocomial infection [30, 31].
In the current study, the difference in the outcome of early pneumonia between the two groups was initially not significant, but with substantial heterogeneity. To reduce substantial heterogeneity and also make the result more robust, sensitivity analysis was employed in accordance with the Cochrane handbook for systematic reviews of interventions . After the significant heterogeneity had disappeared after sensitivity analysis, a significantly better outcome in the antibiotic group could be proven [RR = 0.64, 95% CI (0.42, 0.96), p = 0.03], which is different from a previously published systematic review by Liu et al . This may be due to the different inclusion criteria. Compared with Liu et al`s publication, two studies were not included in the present meta-analysis. One is a non-English-language study . The other did not provide efficiency data to analyze the effects of preventive antibiotics treatment on post-stroke infections and was therefore excluded . Furthermore, we included one more eligible RCT study . A special focus on this is warranted for future research.
Pneumonia is regarded as a predictor of poor outcome and mortality in patients after acute stroke. Several studies have reported that the occurrence of infections after stroke correlates with poor functional outcome and mortality [10, 31]. However, in the present meta-analysis, despite the significant reduction in early and late infection rate through preventive antibiotics, this advantage did not eventually translate to a more favorable outcome or a reduction of early or late mortality. This might support the findings of Vargas et al (2006) who showed that infections after stroke if treated promptly might not be an independent outcome factor, but merely an indicator of the stroke’s severity .
In previous experimental stroke studies, two antibiotics have been shown to be effective on functional outcome. One is minocycline, which seems to have several anti-inflammatory effects, reduces microglial activation, inhibits apoptotic cell death, and has a favorable effect on outcome in experimental stroke studies . However, in a subsequent human research and meta-analysis, preventive intravenous minocycline did not affect the functional outcome in patients after acute stroke . Another potential neuroprotective antibiotic is ceftriaxone, which has been reported reducing mortality and neurological deficits in a rat model of ischemic stroke . In a large clinical trial with 2550 participants after acute stroke however this could not be confirmed for humans. Interestingly, in another randomized controlled trial, prophylactic administration of mezlocillin plus sulbactam was found to be likely associated with a better clinical outcome . However, this result needs to be interpreted with caution because of a relatively small sample of 60 patients.
Correlation between the length of hospital stay and the administration of prophylactic antibiotic treatment had been described in two of the six included articles with a total of 3755 participants [19, 20]. Westendorp et al. (2015) clarified the length of hospital stay did not differ between two groups. Kalra et al. (2015) even found that patients in the antibiotics group had longer stays in hospital than control patients, although this difference was not significant. This indicates that the administration of prophylactic antibiotics may not shorten the hospital stay.
Several factors may bias the findings of the current study. Five studies used preventive antibiotic therapy that covered the common causative organisms in post-stroke infections [11, 13, 15, 19, 20]. However, minocycline, a kind of antibiotic with inadequate microbiological coverage, was used in order to investigate a possible neuroprotective effect in one study , which might have weakened the effect of preventive antibiotics on infection and even outcome. Different definitions for the diagnosis of infection were employed in all these six studies [11, 13, 15, 19, 20, 24], and less strict definitions might underestimate the number of infection events, especially in studies with an open-label design .
The mortality rates were very low in three included studies [15, 20, 24], ranging from 4% to 13%, which is less than previously reported mortality rates of between 15% to 25% in patients with acute stroke . In one of these studies, patients with a life expectancy of less than 90 days were excluded , which obviously could skew reported results of mortality. The selection of patients with different extents of stroke is also a major variable in these studies.
The different effect of preventive antibiotic therapy in patients with a mild stroke versus a severe stroke is hard to grasp. Only two of these six included studies were double-blinded [11, 13], three of the other four studies used an open-label design [19, 20, 24], and in the remaining one study, although infection was assessed blindly, other outcomes were not . Once a medical treatment differs between two groups, knowledge of the intervention in a trial might affect outcome. Although case fatality, a hard endpoint, is not very likely to be influenced, it could have affected the functional score on the mRS, which is a less objective endpoint. Meanwhile, a subjective bias might have occurred by unblinded diagnosis of infection rate, because infections could be diagnosed more easily in the control group, or less easily in the group treated with preventive antibiotic.
Among the six studies included, only two reported the figures of late infections, which limited further analysis on the effect of preventive antibiotic on late infections. Due to the fact that late infections are more often caused by nosocomial multidrug-resistant bacteria, they are probably more difficult to cure. If late infections and overall infections, not just early ones, can be further analyzed, this may help address this issue. This is a shortcoming of the current meta-analysis. Future studies are warranted to further focus on this specific issue.
Additionally, bias might be introduced when participants are withdrawn after randomization. Excluding participants because of an inability to complete the course due to side effects from antibiotics might produce bias in favor of the group treated with preventive antibiotic.
The present meta-analysis of 6 randomized controlled trials including 4125 patients shows that although preventive antibiotic treatment did reduce the occurrence of early overall infections, early pneumonia and early UTI in adults with acute stroke, this advantage was not eventually translated to a shortened hospital stay, favorable outcome and reduction in mortality. Preventive antibiotics are unlikely to have an effect on favorable outcome or mortality in patients with acute stroke. Future studies are warranted to further focus on the late infections and overall incidence of infections, and identify any subgroup of stroke patients who might benefit from preventive antibiotic treatment.
S1 File. Definition of infection in each study.
- 1. Castillo J, Dávalos A, Marrugat J, Noya M. Timing for fever-related brain damage in acute ischemic stroke. Stroke. 1998;29(12):2455–60. pmid:9836750
- 2. Chamorro Á, Meisel A, Planas AM, Urra X, Van De Beek D, Veltkamp R. The immunology of acute stroke. Nat Rev Neurol. 2012;8(7):401–10. pmid:22664787
- 3. Davenport R, Dennis M, Wellwood I, Warlow CP. Complications after acute stroke. Stroke. 1996;27(3):415–20. pmid:8610305
- 4. Grau AJ, Buggle F, Schnitzler P, Spiel M, Lichy C, Hacke W. Fever and infection early after ischemic stroke. J Neurol Sci. 1999;171(2):115–20. pmid:10581377
- 5. Johnston KC, Li JY, Lyden PD, Hanson SK, Feasby TE, Adams RJ, et al. Medical and neurological complications of ischemic stroke experience from the RANTTAS trial. Stroke. 1998;29(2):447–53. pmid:9472888
- 6. Langhorne P, Stott D, Robertson L, MacDonald J, Jones L, McAlpine C, et al. Medical complications after stroke a multicenter study. Stroke. 2000;31(6):1223–9. pmid:10835436
- 7. Weimar C, Roth MP, Zillessen G, Glahn J, rg o, Wimmer ML, et al. Complications following acute ischemic stroke. European neurology. 2002;48(3):133–40. pmid:12373029
- 8. Finlayson O, Kapral M, Hall R, Asllani E, Selchen D, Saposnik G. Risk factors, inpatient care, and outcomes of pneumonia after ischemic stroke. Neurology. 2011;77(14):1338–45. pmid:21940613
- 9. Nakagawa T, Arai H, Kikuchi R, Manabe K, Sasaki H. High incidence of pneumonia in elderly patients with basal ganglia infarction. Arch Intern Med. 1997;157(3):321–4. pmid:9040300
- 10. Popović N, Stefanović-Budimkić M, Mitrović N, Urošević A, Milošević B, Pelemiš M, et al. The frequency of poststroke infections and their impact on early stroke outcome. J Stroke Cerebrovasc Dis. 2013;22(4):424–9. pmid:23540255
- 11. Chamorro A, Horcajada J, Obach V, Vargas M, Revilla M, Torres F, et al. The early systemic prophylaxis of infection after stroke study a randomized clinical trial. Stroke. 2005;36(7):1495–500. pmid:15961713
- 12. Vargas M, Horcajada JP, Obach V, Revilla M, Cervera Á, Torres F, et al. Clinical Consequences of Infection in Patients With Acute Stroke Is It Prime Time for Further Antibiotic Trials? Stroke. 2006;37(2):461–5. pmid:16385093
- 13. Harms H, Prass K, Meisel C, Klehmet J, Rogge W, Drenckhahn C, et al. Preventive antibacterial therapy in acute ischemic stroke: a randomized controlled trial. PloS one. 2008;3(5):e2158. pmid:18478129
- 14. Lampl Y, Boaz M, Gilad R, Lorberboym M, Dabby R, Rapoport A, et al. Minocycline treatment in acute stroke An open-label, evaluator-blinded study. Neurology. 2007;69(14):1404–10. pmid:17909152
- 15. Schwarz S, Al-Shajlawi F, Sick C, Meairs S, Hennerici MG. Effects of Prophylactic Antibiotic Therapy With Mezlocillin Plus Sulbactam on the Incidence and Height of Fever After Severe Acute Ischemic Stroke The Mannheim Infection in Stroke Study (MISS). Stroke. 2008;39(4):1220–7. pmid:18309164
- 16. Westendorp WF, Vermeij JD, Vermeij F, Den Hertog HM, Dippel DW, van de Beek D, et al. Antibiotic therapy for preventing infections in patients with acute stroke. Cochrane Database Syst Rev. 2012;1:CD008530. pmid:22258987
- 17. Conditions NCCfC and London RCoPo. Stroke: national clinical guideline for diagnosis and initial management of acute stroke and transient ischemic attack (TIA). 2008. http://www.rcplondon.ac.uk/pubs/brochure.aspx?e=254 (accessed April 1, 2010).
- 18. Liu L, Xiong XY, Zhang Q, Fan XT, Yang QW. The Efficacy of Prophylactic Antibiotics on Post-Stroke Infections: An Updated Systematic Review and Meta-Analysis. Sci Rep. 2016;6:36656. pmid:27841284
- 19. Kalra L, Irshad S, Hodsoll J, Simpson M, Gulliford M, Smithard D, et al. Prophylactic antibiotics after acute stroke for reducing pneumonia in patients with dysphagia (STROKE-INF): a prospective, cluster-randomised, open-label, masked endpoint, controlled clinical trial. Lancet. 2015;386(10006):1835–44. pmid:26343840
- 20. Westendorp WF, Vermeij J-D, Zock E, Hooijenga IJ, Kruyt ND, Bosboom HJ, et al. The Preventive Antibiotics in Stroke Study (PASS): a pragmatic randomised open-label masked endpoint clinical trial. Lancet. 2015;385(9977):1519–26. pmid:25612858
- 21. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med. 2009;151(4):W-65-W-94.
- 22. Higgins J and Green S. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1. 0. The Cochrane Collaboration. 2011;5 www.cochrane-handbook.org (accessed 2012 June 6).
- 23. Deeks JJ, Higgins J, Altman DG. Analysing data and undertaking meta-analyses. Cochrane handbook for systematic reviews of interventions: Cochrane book series. 2008:243–96.
- 24. Kohler E, Prentice DA, Bates TR, Hankey GJ, Claxton A, van Heerden J, et al. Intravenous Minocycline in Acute Stroke A Randomized, Controlled Pilot Study and Meta-analysis. Stroke. 2013;44(9):2493–9. pmid:23868273
- 25. Emsley HC, Hopkins SJ. Acute ischaemic stroke and infection: recent and emerging concepts. Lancet Neurol. 2008;7(4):341–53. pmid:18339349
- 26. Indredavik B, Rohweder G, Naalsund E, Lydersen S. Medical complications in a comprehensive stroke unit and an early supported discharge service. Stroke. 2008;39(2):414–20. pmid:18096834
- 27. Ionita CC, Siddiqui AH, Levy EI, Hopkins LN, Snyder KV, Gibbons KJ. Acute ischemic stroke and infections. J Stroke Cerebrovasc Dis. 2011;20(1):1–9. pmid:20538486
- 28. Masiero S, Pierobon R, Previato C, Gomiero E. Pneumonia in stroke patients with oropharyngeal dysphagia: a six-month follow-up study. Neurol Sci. 2008;29(3):139–45. pmid:18612760
- 29. Prass K, Braun JS, Dirnagl U, Meisel C, Meisel A. Stroke propagates bacterial aspiration to pneumonia in a model of cerebral ischemia. Stroke. 2006;37(10):2607–12. pmid:16946159
- 30. Ros L, García M, Prat J, González C, Gimeno C, Albert A, et al. Predictors of nosocomial infection in acute stroke. Relation with morbimortality and outcome. Med Clin (Barc). 2007;128(12):441–7.
- 31. Sellars C, Bowie L, Bagg J, Sweeney MP, Miller H, Tilston J, et al. Risk factors for chest infection in acute stroke a prospective cohort study. Stroke. 2007;38(8):2284–91. pmid:17569875
- 32. De Falco FA, Santangelo R, Majello L, Angelone P. Antimicrobial prophylaxis in the management of ischemic stroke. Rivista di Neurobiologia. 1998;44(1):63–7.
- 33. Lampl Y, Boaz M, Gilad R, Lorberboym M, Dabby R, Rapoport A, et al. Minocycline treatment in acute stroke An open-label, evaluator-blinded study. Neurology. 2007;69(14):1404–10. pmid:17909152
- 34. Yong VW, Wells J, Giuliani F, Casha S, Power C, Metz LM. The promise of minocycline in neurology. Lancet Neurol. 2004;3(12):744–51. pmid:15556807
- 35. Thöne-Reineke C, Neumann C, Namsolleck P, Schmerbach K, Krikov M, Schefe JH, et al. The β-lactam antibiotic, ceftriaxone, dramatically improves survival, increases glutamate uptake and induces neurotrophins in stroke. J Hypertens. 2008;26(12):2426–35. pmid:19008722
- 36. Kishore AK, Vail A, Chamorro A, Garau J, Hopkins SJ, Di Napoli M, et al. How is pneumonia diagnosed in clinical stroke research? A systematic review and meta-analysis. Stroke. 2015;46(5):1202–9. pmid:25858238
- 37. van der Worp HB, van Gijn J. Acute ischemic stroke. N Engl J Med. 2007;357(6):572–9. pmid:17687132