Little is known about risk factors for severe outcomes in patients infected with H5N1 and no systematic review has been conducted. Understanding risk factors is an important step for prioritizing prophylaxis or treatment in the event of a pandemic.
To systematically evaluate risk factors for severe outcomes in patients with avian influenza H5N1 infection.
Eligibility criteria for selecting studies
Observational studies of any design published in English, French, Spanish, German or Korean that reported on risk factor-outcome combinations of interest in participants with confirmed H5N1 infections. Outcomes considered included death, ventilator support, hospital and ICU admission, pneumonia, and composite outcomes.
We identified 20 studies reporting on 999 patients infected with H5N1. The majority of studies (n = 14, 70%) were at intermediate risk of bias, i.e. 4–6 points on the NOS. Females were at increased risk of death (OR 1.75, 95% CI 1.27–2.44), while young age, in particular <5 years of age (OR 0.44, 95% CI 0.25–0.79 for death), was protective. Data on traditional risk factors was scarce and requires further studies. Another major limitation in the published literature was lack of adjustment for confounders.
Citation: Mertz D, Kim TH, Johnstone J, Lam P-P, Science M, Kuster SP, et al. (2014) Populations at Risk for Severe or Complicated Avian Influenza H5N1: A Systematic Review and Meta-Analysis. PLoS ONE 9(3): e89697. https://doi.org/10.1371/journal.pone.0089697
Editor: Paul Robert Cleary, Health Protection Agency, United Kingdom
Received: December 2, 2013; Accepted: January 21, 2014; Published: March 6, 2014
Copyright: © 2014 Mertz 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.
Funding: The World Health Organization (WHO) funded the study. The protocol was reviewed by the WHO and the WHO's suggestions were incorporated into the protocol. DM was partly supported by a research scholarship from the Swiss National Science Foundation (PASMP3-132571) and the Lichtenstein-Stiftung, and is a recipient of a Research Early Career Award from Hamilton Health Sciences Foundation (Jack Hirsh Fellowship); ML holds the Michael G. DeGroote Chair in Infectious Diseases at McMaster University; JJ receives salary support from the Canadian Institutes of Health Research. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: Mark Loeb has been a paid consultant for GSK, Novartis, and Sanofi Pasteur (vaccine manufacturers). There are no other competing interests to be declared. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.
Outbreaks of highly pathogenic avian influenza H5N1 infections in poultry were first reported from Guangdong, China in 1996 . Since a cluster in Hong Kong in 1997, transmission to humans has been intermittently reported and, as of April 2013, a total of 628 confirmed cases with 374 deaths of H5N1 infected patients (59.6% case fatality rate) have been reported to the World Health Organization (WHO) from 15 countries .
Understanding risk factors for severe outcomes is an important step in order to prioritize prophylaxis or treatment in the event of a pandemic. As part of a systematic review sponsored by the WHO, we aimed to identify risk factors for severe outcomes or complications of influenza infections. We report our findings specific to highly pathogenic avian influenza H5N1.
Materials and Methods
The methods were reported in detail previously . Briefly, we included studies reporting on at least one risk factor-outcome combination of interest irrespective of the study design. Only single-case reports were excluded, but case series reporting on patients with and without a specific outcome for a particular risk factor of interest were included. Articles based on secondary analyses of previously published data and review articles were not eligible. Outcomes of interest included community-acquired pneumonia, mortality, hospitalization, intensive care unit (ICU) admission, need for ventilator support, and composite outcomes. Studies in English, French, German, Spanish, and Korean were included.
We searched MEDLINE, EMBASE, CINAHL, Global Health, and the Cochrane Central Register of Controlled Trials (CENTRAL) up to March 29 of 2011 and reference lists of identified articles and review articles. Screening and data abstraction were conducted independently and in duplicate by pairs of reviewers using piloted and standardized forms. Two reviewers assessed the risk of bias independently using the Newcastle-Ottawa scale (NOS) .
A random-effects model was used to calculate summary estimates using Review Manager 5.0 (Cochrane Collaboration) . Risk estimates are reported as odds ratio (OR) with 95% confidence intervals (CI). If there was a potential overlap in study populations across studies in a meta-analysis, studies were excluded in a sensitivity analysis in order to minimize any effect of such overlap. The I2 statistics was used to evaluate heterogeneity . Two a priori defined subgroup analysis based on the place of enrollment (community versus hospital versus ICU population) and risk of bias in included studies were conducted in the presence of significant heterogeneity defined by I2>60% . Funnel plots were visually interpreted to assess publication bias but no formal statistical tests were conducted due to the small number of studies for the vast majority of risk factor-outcome combinations.
The World Health Organization (WHO) funded the study. The protocol was reviewed by the WHO and the WHO's suggestion were incorporated into the protocol.
Out of 47,874 titles and abstracts screened, 20 studies , , , , , , , , , , , , , , , , , , ,  that included 999 patients reported risk factors for severe outcomes for H5N1 avian influenza (Figure 1). All studies were cohort studies and published in English between 1998 and 2010. Evidence of H5N1 influenza infection was by laboratory confirmation in all studies. Only two studies reported on cases outside of Asia.
As assessed by the Newcastle-Ottawa scale, the majority of studies were at intermediate risk of bias (i.e. 4–6 points; n = 14, 70%) while 4 (20%) were at highest risk of bias (i.e. 3 or less points) and 2 (10%) at lowest risk of bias (i.e. 7–9 points). Although the adjusted odds ratios were reported in only one study, the use of adjusted odds ratios instead of the crude odds ratios would not have changed our assessment of the risk factors.
Female sex was associated with higher all-cause mortality (OR 1.75, 95% 1.27–2.44) (Figure 2, Table 1). We also found a non-significant trend towards higher rates of pneumonia (OR 3.03, 95% CI 0.91–10.00), ICU admission (OR 3.70, 95% CI 0.33–50.00), and need for ventilator support (OR 2.94, 95% CI 0.43–20.00) in women. Because overlap in study populations was likely across some studies for mortality and pneumonia outcomes, a sensitivity analysis was conducted excluding studies with a potential overlap. The findings were similar (Table 1).
We found no evidence that older age was a risk factor for more severe outcomes when comparing >65 years of age to non-elderly adults (OR for all-cause mortality 1.04, 95% CI 0.12–9.08). Children <18 years of age were at lower risk of developing pneumonia (OR 0.16, 95% CI 0.03–0.86) when compared to adult patients. There was a non-significant tendency towards lower mortality rates (0.48, 95% CI 0.13–1.79) and lower risk of pneumonia (OR 0.12, 95% CI 0.01–1.03) when compared to non-elderly adults. We found that children of <5 years of age were at lower risk of death when compared to older children 5–18 years of age (OR 0.44, 0.25–0.79; Figure 3) with a non-significant trend towards a lower likelihood of developing pneumonia or need ventilator support.
Only two studies reported on the presence of co-morbidities as potential traditional risk factors. Although no significant differences were found, there was a non-significant tendency towards increased risk of pneumonia in the presence of co-morbidities (OR 1.71, 95% CI 0.35–8.36). No significant differences in outcomes were found for smoking.
The only comparison with significant heterogeneity was for all-cause mortality when comparing paediatric patients to non-elderly adults (I2 = 73%). A subgroup analysis excluding studies at highest risk of bias (Newcastle-Ottawa scale of 3 or lower) reduced the I2 to 0% with a smaller overall effect (OR 0.90, 95% CI 0.51–1.57 versus OR 0.48, 95% CI 0.13–1.79) in the primary analysis.
We found that female sex was associated with worse outcomes following H5N1 infection and that children, particularly those <5 years old, had better outcomes. Data on underlying co-morbidities as risk factors for severe outcomes were scarce and therefore there is no evidence demonstrating that co-morbidities affect the outcome in patients infected with H5N1.
At the time of our literature search, 539 human cases with H5N1 infections from 15 countries had been reported to WHO . In the 20 studies identified in our systematic review, a total of 999 cases were reported. While it seems unlikely that a large number of confirmed H5N1 infections have not been reported to WHO , the excess in cases in combination with identical time periods and regions reported in included studies indicates overlap in patient populations reported in the literature. In particular, there was an overlap between two large WHO reports (Appendix A, references 11 and 13) and studies published by local investigators. In a sensitivity analysis we excluded studies that appeared to have included some participants reported in another included study, the results were in keeping with the primary analyses. Since the overlap in participants between studies was never 100%, the full dataset was presented as the primary analysis.
Female sex was associated with a higher mortality risk as well as with a non-significant trend for other severe outcomes. Our findings are in line with a report from Egypt where the H5N1 case fatality rate was 49% in females compared to 8% in males . Of note, the overall mortality rate was lower in Egypt as compared to mortality rates reported from other countries (34 versus 66%) . A higher mortality in females was also found in a review of 294 cases occurring between 2006 and 2010 published by Fiebig et al. . Women infected with H5N1 were older  and most of the affected women were housewives taking care of poultry . In contrast, the males were largely pre-school boys that got exposed while playing outdoors . An exposure to a greater inoculum is one possibility for the worse outcomes in women. Another is that the sex differences were confounded by age: females were likely exposed at an older age than boys, and it was shown that younger age, in particular <5 years of age, was associated with a lower risk for severe outcomes. This is in keeping with the findings in the above mentioned study by Fiebig et al. There, the authors found an even more pronounced difference with five to six times lower odds for death for children up to 9 years of age as compared to a number of other age groups .
In contrast to non-H5N1 influenza infections where older age was shown to be a risk factor for severe outcomes for seasonal influenza as well as during the 2009 H1N1 pandemic , this does not seem to be true for H5N1 infections. One can hypothesize that previous exposure to other subtypes may have offered protection against H5N1 .
One limitations of this systematic review is the potential for reporting bias. In particular asymptomatic patients or patients with minor symptoms were likely underrepresented in the cases reported. In fact, a 1 to 2% infection rate in those exposed was found in serological studies which would result in much higher number of cases than reported , which in turn is regarded to be an overestimate by others . Adjusted risk estimates also were not available. Thus, modifying risk factors such as differing time to oseltamivir treatment or the availability of health care may have biased outcomes , . Also, we did not consider articles in Chinese or other languages spoken in the regions where H5N1 is endemic. However, we did not find any relevant articles on H5N1 when screening titles and abstracts of the 50 Chinese articles that were excluded. Among the remaining 220 articles that were excluded due to language, we only identified three review articles on H5N1 when searching for the key words “avian” or “H5N1” in the title. Thus, we believe that we did not miss any relevant articles on Avian Influenza in the databases that we searched due to restrictions on language. Finally, sample sizes for some rarely reported risk factor-outcome comparisons were small explaining the low precision and likely the lack of statistical significance for some of these trends that we noted.
In summary, we found evidence that females were at increased risk for complications following H5N1 infection while young age protected against severe outcomes. Research on traditional risk factors is scarce and is required.
Conceived and designed the experiments: DM THK JJ NB ML. Performed the experiments: DM THK JJ PL MS SPK SAF DT EF. Analyzed the data: DM THK JJ PL MS SPK SAF DT EF. Contributed reagents/materials/analysis tools: NB. Wrote the paper: DM THK JJ ML.
- 1. Yin J, Liu S, Zhu Y (2013) An overview of the highly pathogenic H5N1 influenza virus. Virologica Sinica 28: 3–15.
- 2. World Health Organization (2013) Cumulative number of confirmed human cases of avian influenza A(H5N1) reported to WHO. http://www.who.int/influenza/human_animal_interface/EN_GIP_20130426CumulativeNumberH5N1cases.pdf, accessed 2013 Dec 1.
- 3. Mertz D, Kim TH, Johnstone J, Lam PP, Science M, et al. (2013) Populations at risk for severe or complicated influenza illness: systematic review and meta-analysis. BMJ 347: f5061.
- 4. Wells GA, Shea B, O'Connell D, Peterson J, Welch V, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp, accessed Dec 1 2013.
- 5. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7: 177–188.
- 6. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ 327: 557–560.
- 7. The Cochrane Collaboration (2011) Cochrane Handbook for Systematic Reviews of Interventions. www.cochrane-handbook.org, accessed 2013 Dec 1.
- 8. World Health Organization (2006) Epidemiology of WHO-confirmed human cases of avian influenza A(H5N1) infection. Wkly Epidemiol Rec 81: 249–257.
- 9. World Health Organization (2007) Update: WHO-confirmed human cases of avian influenza A(H5N1) infection, 25 November 2003–24 November 2006. Wkly Epidemiol Rec 82: 41–47.
- 10. Areechokchai D, Jiraphongsa C, Laosiritaworn C, Hanshaoworakul W, O'Reilly M (2006) Investigation of avian influenza (H5N1) outbreak in humans – Thailand, 2004. MMWR: Morbidity & Mortality Weekly Report 6.
- 11. Chotpitayasunondh T, Ungchusak K, Hanshaoworakul W, Chunsuthiwat S, Sawanpanyalert P, et al. (2005) Human disease from influenza A (H5N1), Thailand, 2004. Emerging Infectious Diseases 11: 201–209.
- 12. Fasina FO, Ifende VI, Ajibade AA (2010) Avian influenza A(H5N1) in humans: lessons from Egypt. Euro Surveillance: Bulletin Europeen sur les Maladies Transmissibles = European Communicable Disease Bulletin 15: 19473.
- 13. Giriputro S, Agus R, Sulastri S, Murniati D, Darwis F, et al. (2008) Clinical and epidemiological features of patients with confirmed avian influenza presenting to Sulianti Saroso Infectious Diseases Hospital, Indonesia, 2005–2007. Annals of the Academy of Medicine, Singapore 37: 454–457.
- 14. Kandun IN, Tresnaningsih E, Purba WH, Lee V, Samaan G, et al. (2008) Factors associated with case fatality of human H5N1 virus infections in Indonesia: a case series. Lancet 372: 744–749.
- 15. Kandun IN, Wibisono H, Sedyaningsih ER, Yusharmen, Hadisoedarsuno W, et al. (2006) Three Indonesian clusters of H5N1 virus infection in 2005. New England Journal of Medicine 355: 2186–2194.
- 16. Liem NT, Tung CV, Hien ND, Hien TT, Chau NQ, et al. (2009) Clinical features of human influenza A (H5N1) infection in Vietnam: 2004–2006. Clinical Infectious Diseases 48: 1639–1646.
- 17. Olsen SJ, Laosiritaworn Y, Pattanasin S, Prapasiri P, Dowell SF (2005) Poultry-handling practices during avian influenza outbreak, Thailand. Emerging Infectious Diseases 11: 1601–1603.
- 18. Oner AF, Bay A, Arslan S, Akdeniz H, Sahin HA, et al. (2006) Avian influenza A (H5N1) infection in eastern Turkey in 2006. New England Journal of Medicine 355: 2179–2185.
- 19. Peiris JSM, Yu WC, Leung CW, Cheung CY, Ng WF, et al. (2004) Re-emergence of fatal human influenza A subtype H5N1 disease. Lancet 363: 617–619.
- 20. Pollack CV Jr, Kam CW, Mak YK (1998) Update: isolation of avian influenza A(H5N1) viruses from human beings–Hong Kong, 1997–1998. Annals of Emergency Medicine 31: 647–649.
- 21. Sedyaningsih ER, Isfandari S, Setiawaty V, Rifati L, Harun S, et al. (2007) Epidemiology of cases of H5N1 virus infection in Indonesia, July 2005–June 2006. Journal of Infectious Diseases 196: 522–527.
- 22. Soepandi PZ, Burhan E, Mangunnegoro H, Nawas A, Aditama TY, et al. (2010) Clinical course of avian influenza A(H5N1) in patients at the Persahabatan Hospital, Jakarta, Indonesia, 2005–2008. Chest 138: 665–673.
- 23. Tam JS (2002) Influenza A (H5N1) in Hong Kong: an overview. Vaccine 20 Suppl 2: S77–81.
- 24. Tran TH, Nguyen TL, Nguyen TD, Luong TS, Pham PM, et al. (2004) Avian influenza A (H5N1) in 10 patients in Vietnam. New England Journal of Medicine 350: 1179–1188.
- 25. Yu H, Feng Z, Zhang X, Xiang N, Huai Y, et al. (2007) Human influenza A (H5N1) cases, urban areas of People's Republic of China, 2005–2006. Emerging Infectious Diseases 13: 1061–1064.
- 26. Yu H, Gao Z, Feng Z, Shu Y, Xiang N, et al. (2008) Clinical characteristics of 26 human cases of highly pathogenic avian influenza A (H5N1) virus infection in China. PLoS ONE [Electronic Resource] 3: e2985.
- 27. Yuen KY, Chan PK, Peiris M, Tsang DN, Que TL, et al. (1998) Clinical features and rapid viral diagnosis of human disease associated with avian influenza A H5N1 virus. Lancet 351: 467–471.
- 28. World Health Organization (2011) Influenza at the human-animal interface. Summary and assessment of March 2011 events. http://www.who.int/influenza/human_animal_interface/Influenza_Summary_IRA_HA_interface_07April11.pdf, Accessed 2013 Dec 1.
- 29. Fiebig L, Soyka J, Buda S, Buchholz U, Dehnert M, et al. (2011) Avian influenza A(H5N1) in humans: new insights from a line list of World Health Organization confirmed cases, September 2006 to August 2010. Euro surveillance: bulletin Europeen sur les maladies transmissibles = European communicable disease bulletin 16.
- 30. Dudley JP (2009) Age-specific infection and death rates for human A(H5N1) avian influenza in Egypt. Euro surveillance : bulletin Europeen sur les maladies transmissibles = European communicable disease bulletin 14.
- 31. Chen JM, Chen JW, Dai JJ, Sun YX (2007) A survey of human cases of H5N1 avian influenza reported by the WHO before June 2006 for infection control. American journal of infection control 35: 467–469.
- 32. Buricchi F, Bardelli M, Malzone C, Capecchi B, Nicolay U, et al. (2013) Impact of preexisting memory to seasonal A/H1N1 influenza virus on the immune response following vaccination against avian A/H5N1 virus. European journal of immunology 43: 641–648.
- 33. Wang TT, Parides MK, Palese P (2012) Seroevidence for H5N1 influenza infections in humans: meta-analysis. Science 335: 1463.
- 34. Van Kerkhove MD, Riley S, Lipsitch M, Guan Y, Monto AS, et al. (2012) Comment on “Seroevidence for H5N1 influenza infections in humans: meta-analysis”. Science 336: 1506 author reply 1506.
- 35. Adisasmito W, Chan PK, Lee N, Oner AF, Gasimov V, et al. (2010) Effectiveness of antiviral treatment in human influenza A(H5N1) infections: analysis of a Global Patient Registry. The Journal of infectious diseases 202: 1154–1160.