Abstract
Background
Climate factors have been shown to be associated with spontaneous musculoskeletal and some surgical site infections with increased rates of infection during warmer periods. To date, little research has been performed to determine if this phenomenon is associated with differences in the risk of revision for prosthetic joint infection (PJI) in primary TKA.
Questions/purposes
(1) Does the rate of revision for early PJI within the first year after primary TKA differ between tropical and nontropical regions? (2) Is there a seasonal variation in the rate of revision for PJI? (3) Is the geographic and seasonal variation (if present) associated with the sex, age, and/or American Society of Anesthesiologists (ASA) grade of the patient?
Methods
All 219,983 primary TKAs performed for osteoarthritis over a 5-year period (2011-2015) in the Australian Orthopaedic Association National Joint Replacement Registry were examined based on the month of the primary procedure to determine the rate of revision for PJI within 12 months. The data were analyzed to determine the differences in the risk of revision for PJI based on geographic region and season of the primary procedure adjusting for sex, age, and ASA grade of the patient.
Results
The early revision rate for PJI was higher in the tropical compared with the nontropical region of Australia (0.73% versus 0.37%; odds ratio [OR], 1.87; 95% confidence interval [CI], 1.44-2.42; p < 0.001). The tropical region of Australia demonstrated a seasonal variation in the rate of revision for PJI with a higher rate during the warmer monsoon wet season of summer and fall (summer/fall 0.98% versus winter/spring 0.51%; OR, 1.88; 95% CI, 1.12-3.16; p = 0.02). A seasonal variation was not seen in the nontropical region (OR, 1.03; 95% CI, 0.90-1.19; p = 0.64). The regional and seasonal changes were independent of sex, age, and ASA grade.
Conclusions
Climate factors are associated with the risk of early revision for PJI in patients undergoing primary TKA with rates of such revisions approximately double in tropical regions compared with nontropical regions. Additionally, tropical regions demonstrate a seasonal variation with the risk of PJI doubling during the warmer, monsoonal wet season of summer and fall. These findings should be confirmed in further studies that can better control for possible confounding variables. The mechanism for this phenomenon is not clear, and further research into this subject is also indicated.
Level of Evidence
Level III, therapeutic study.
Introduction
Over the last few decades, the risk factors for the development of prosthetic joint infection (PJI) have become more clearly defined [16], yet one area that has not been fully explored is the association of climate factors on the risk of developing a PJI. There is evidence of a seasonal variation in the rates of musculoskeletal infections such as cellulitis, myositis, and osteomyelitis with an increased incidence during warmer periods [4, 8, 12, 13, 15, 17, 19, 20]. A seasonal variation has also been shown in the rates of surgical site infections after cardiothoracic, spinal, and lower extremity fracture procedures [9, 18, 21]. Finally, a recent study by Anthony et al. [2] has analyzed > 55 million surgical procedures and correlated the risk of surgical site infection (SSI) with climate factors. They found the SSI incidence was highly seasonal with the highest SSI admission rate during the warmest month of August and the lowest in the coldest month of January. The odds of a SSI admission increased by 2.1% per 2.8°C increase in the average monthly temperature with the highest temperature group having an increased odds of 29% compared with the lowest temperature group.
To our knowledge, there is only one published study examining the seasonal variation of arthroplasty infection rates. This small single-surgeon case series demonstrated a higher infection rate for hip and knee arthroplasties performed during the summer months [11]. The current study sought to confirm whether there is an association between climate factors and early revision for PJI after primary TKA by using the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) database.
Specifically, our aim was to determine if climate/seasonal factors played a role in the risk of early (< 1 year) revision for PJI after primary TKA. We asked: (1) Does the rate of revision for early PJI within the first year after primary TKA differ between tropical and nontropical regions? (2) Is there a seasonal variation in the rate of early revision for PJI? (3) Is the geographic and seasonal variation (if present) influenced by the sex, age, and/or American Society of Anesthesiologists (ASA) grade of the patient?
Patients and Methods
All primary TKAs that were performed for osteoarthritis over a 5-year period (January 2011 to December 2015 to reflect current practice) in Australia and registered in the AOANJRR were included in this study. The AOANJRR captures > 98% of all joint replacement procedures performed in Australia. A total of 219,983 primary TKAs were performed during the study period, with 8344 procedures in tropical regions and 211,639 in nontropical regions.
The relevant data obtained from the AOANJRR included baseline patient demographics, ASA grade, body mass index (BMI), initial diagnosis, primary operation date, location, and details regarding any subsequent revision procedures, including the indication for revision. The AOANJRR began recording ASA grade in 2012 and BMI in 2015. The subset of the study population with BMI data was thus too small for analysis. ASA grade comorbidity data were available for a subset of 125,037 primary TKAs (4375 tropical, 120,662 nontropical, 59% of total study cohort) and have been included in a secondary analysis.
The primary outcome for this study was early revision for infection within 12 months of the index procedure. The registry records a revision procedure if any component is removed, exchanged, or inserted, eg, polyethylene liner exchange. The registry does not capture PJIs that are treated without insertion, removal, or exchange of a prosthesis component. The date of the index procedure, not the date of the revision for infection, was used for the seasonal analysis.
The study cohort was separated into two geographic regions within Australia (tropical and nontropical). Tropical regions of Australia included the cities above the tropic of Capricorn (Rockhampton, Mackay, Townsville, Cairns, and Darwin). The tropic of Capricorn (26° south) is the dividing latitude between the Southern Temperate Zone to the south and the Tropics to the north. The tropics of Australia generally have two seasonal phases (dry and wet seasons) that are linked to wind directions. In the winter phase (winter/spring), easterly tradewinds bring dry and cooler conditions. In the summer phase (summer/fall), westerly winds bring warm, moist air and sustained rainy conditions.
To identify if socioeconomic factors (as a potential confounder) differed between the two regions, the 2011 Australian Bureau of Statistics (ABS) Index of Relative Socio-economic Advantage and Disadvantage (IRSAD) for each tropical city was examined (Table 1). The IRSAD summarizes the economic and social conditions of people and households within an area and ranks them within the Australian percentile with a lower percentile representing a socioeconomic disadvantage. All cities within the tropical region of Australia have an IRSAD at or above the 50th percentile of the country.
Table 1.
Index of relative socioeconomic advantage and disadvantage for tropical cities
A generalized linear model, assuming a binomial distribution with a logit link, was used to determine if there was a difference in the probability of revision for infection within 1 year between tropical and nontropical regions, seasons, age, and sex of the patient. A secondary analysis of available data was performed adjusting for the ASA grade. Odd ratios are presented with 95% confidence intervals. Statistical significance was set at 5%. Statistical analysis was performed using SAS software, Version 9.4 (SAS Institute Inc, Cary, NC, USA). This study was approved and conducted according to the ethical standards of the AOANJRR.
Results
There was a higher proportion of revisions for early PJI in the tropical region compared with the nontropical region after adjusting for age and sex (0.73% versus 0.37%; odds ratio [OR], 1.87; 95% confidence interval [CI], 1.44-2.42; p < 0.001).
When examining for a seasonal variation between the dry and wet seasons (adjusting for age and sex), only the tropical region of Australia demonstrated a higher proportion of revisions for early PJI in the wet season (summer/fall 0.98% versus winter/spring 0.51%; OR, 1.88; 95% CI, 1.12-3.16; p = 0.02). There was no seasonal variation in the nontropical region of Australia (OR, 1.03; 95% CI, 0.90-1.19; p = 0.64).
Overall, males had a higher rate of revision for PJI than females (adjusting for age) (OR, 2.48; 95% CI, 2.15-2.86; p < 0.001). The interaction between region and season (described previously) was independent of sex and age of the patient.
The secondary analysis adjusting for ASA grade did not alter the previously mentioned study findings with the differences in PJI rates between regions, seasons, and sex unchanged (tropical versus nontropical PJI rates [adjusting for age, sex, and ASA]: 0.64% versus 0.34%; OR, 1.78; 95% CI, 1.21-2.62; p = 0.03; tropical seasonal variation [adjusting for age, sex, and ASA; summer/fall 0.96% versus winter/spring: 0.4%; OR, 2.42; 95% CI, 1.11-5.25; p = 0.03).
Discussion
Previous studies have reported a seasonal variation in musculoskeletal and SSI rates [2, 6, 8, 9, 11, 13, 15, 17-19], but a phenomenon of an increased risk of early arthroplasty revision and seasonal variation of SSI rates in tropical regions has not previously been described. The important finding of this study is that climate factors appear associated with the risk of early revision for PJI after primary TKA. Compared with nontropical regions, tropical regions have an increased risk of revision and within this region, there is a seasonal variation with increased infection rates during the warmer humid months. The associations seen were independent of any differences in age, sex, or ASA grade.
There are numerous limitations associated with our study. Because the incidence of PJI after primary TKA is expectantly low, it is very difficult to appropriately power and conduct a study on PJI across multiple geographic regions. The utilization of AOANJRR data has allowed us to capture all primary TKAs within a single country to answer our study questions, but registry data have limitations. First, the study findings may not be generalizable to other countries, because perioperative infection prevention practices vary. Within our country, hospitals must report their compliance with national perioperative antibiotic prophylaxis guidelines, which is considered the single most important preventive factor for PJI [1]. Although this process helps to ensure consistency of antibiotic prophylaxis administration across the country, other perioperative prevention methods are not accounted for. For example, the variation of staphylococcal screening and decolonization protocols across hospitals is unknown. In an attempt to reduce the potential confounders of the gradual change in arthroplasty infection prevention strategies over time, we used only the most recent 5-year registry data to reflect modern arthroplasty practice.
Patient confounders between the two study regions are another important limitation. The socioeconomic index did not show that patients in tropical regions are disadvantaged compared with the remainder of the country. Unfortunately, our BMI data were too limited for use in this analysis. The AOANJRR started collecting ASA data during the study period and this information was available for over half of the study cohort. After adjusting for ASA grade, the results of the study were unchanged. Third, the study has examined early PJI by analyzing the revision rate of TKA for PJI. We acknowledge that this analysis only captures cases of PJI that involved at least a change in one component of the prosthesis, for example the polyethylene liner. The number of patients who underwent isolated antibiotic suppressive treatment or débridement without exchange of modular components is unknown. However, we do not expect this to confound the seasonal or geographic variation. The analysis for the seasonal variation of infection rates has been calculated by the month of the index procedure, not the month of revision for infection. This decision was based on the knowledge that the majority of early PJIs are thought to occur from seeding of the joint at the time of the index procedure. We acknowledge that in a small number of cases, the PJI could have occurred at a later date from another source spreading to the joint. Given that the risk and incidence of other potential seeding musculoskeletal infections have been shown to be higher in the summer months, this bias would potentially increase the incidence of PJI rates for the winter cohort, which would only decrease the odds of a difference between the seasons being found. Finally, the division between tropical and nontropical and wet and dry seasons was based on convenience (the true dates of the wet season vary). However, we would consider any misclassification errors to bias the results toward the null.
There was a higher proportion of revisions for early PJI in the tropical region compared with the nontropical region after adjusting for age and sex. The possible explanations for this difference seen between tropical and nontropical regions of Australia could relate to the climatic differences between these two areas. The characteristic climatic difference between the upper tropical zone and the lower nontropical zone of Australia is the large variation in humidity. Tropical regions have a small variation in temperature throughout the year but a large variation in humidity with a peak during the monsoonal wet season (December to May). Comparatively, the nontropical region of Australia has a low and even reversed variation in humidity and a larger variation in temperature. For example, the tropical city of Darwin has a temperature variation of 31.7°C versus 29.6°C and humidity variation of 72% versus 37% between the warmest and coolest months of the year. In comparison, the nontropical city of Melbourne has a temperature variation 24.7°C versus 13.0°C and humidity variation of 47% versus 63% between the warmest and coolest months of the year (data from the Australian Bureau of Meteorology). It is possible that it is the combined relationship of these variables that is associated with the variation of risk of revision for PJI that is seen in tropical regions. The potential association of humidity was first reported in 1985 by Gillespie [8] who mapped out the seasonal variation in the incidence of acute osteomyelitis over a 20-year period from three continents and found it to correlate with periods of raised humidity.
After adjusting for age and sex, we found that only the tropical region of Australia demonstrated an association with seasonal variation in association with revisions for early PJI; there was no seasonal variation in the nontropical region of Australia. Although most of the current literature on this topic supports a seasonal variation of infection rates [2, 6,7,8,9, 11, 13, 15, 17, 18], some studies have not demonstrated such an effect. Two recent single-center studies have failed to demonstrate a seasonal variation of SSI rates [10, 22]. These studies were undertaken at hospitals in nontropical regions and as this current study has shown, seasonal variations in nontropical regions are not associated with the risk of SSI.
In conclusion, this study has shown that climate factors are associated with an increased risk of early revision for PJI after TKA. Tropical regions demonstrated an increased incidence of PJI and a seasonal variation of infection rates with an increased incidence during the warmer wet season. The mechanism by which warm and humid weather increases SSI rates is currently unknown. One of the common hypotheses is that during periods of increased temperature and humidity, the cutaneous environment becomes optimal for the proliferation of commensal organisms such as Staphylococcus aureus. The increased skin bacterial load is thought to contribute to an increased risk of perioperative contamination of the surgical wound. This theory has been reinforced by basic science, laboratory, and clinical studies [3, 5, 14, 23]. Our findings should be confirmed in further studies that can better control for possible confounding variables. Additional research to confirm the mechanism of this phenomenon and find preventive measures for patients in tropical regions is also required.
Acknowledgments
We thank the AOANJRR and the hospitals, orthopaedic surgeons, and patients whose data made this work possible.
Footnotes
Each author certifies that neither he or she, nor any member of his or her immediate family, has funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article. The Australian Orthopaedic Association National Joint Replacement Registry is funded by the Commonwealth of Australia’s Department of Health and Ageing.
All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.
Each author certifies that his or her institution approved the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.
This work was performed at Cairns Hospital, Cairns, Queensland, Australia, in conjunction with the Australian Orthopaedic Association National Joint Replacement Registry.
References
- 1.Alijanipour P, Heller S, Parvizi J. Prevention of periprosthetic joint infection: what are the effective strategies? J Knee Surg. 2014;27:251–258. [DOI] [PubMed] [Google Scholar]
- 2.Anthony CA, Peterson RA, Polgreen LA, Sewell DK, Polgreen PM. The seasonal variability in surgical site infections and the association with warmer weather: a population-based investigation. Infect Control Hosp Epidemiol. 2017;38:809–816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Benjamin E, Reznik A, Williams AL. Mathematical models for conventional and microwave thermal deactivation of Enterococcus faecalis, Staphylococcus aureus and Escherichia coli. Cell Mol Boil (Noisy-le-grand). 2007;53;42–48. [PubMed] [Google Scholar]
- 4.Dillon HC. Impetigo contagiosa: suppurative and non-suppurative complications: I. Clinical, bacteriologic, and epidemiologic characteristics of impetigo. Am J Dis Child. 1968;115:530–541. [DOI] [PubMed] [Google Scholar]
- 5.Duncan WC, McBride ME, Knox JM. Bacterial flora. The role of environmental factors. J Invest Dermatol. 1969;52:479–484. [DOI] [PubMed] [Google Scholar]
- 6.Durkin MJ, Dicks KV, Baker AW, Lewis SS, Moehring RW, Chen LF, Sexton DJ, Anderson DJ. Seasonal variation of common surgical site infections: does season matter? Infect Control Hosp Epidemiol. 2015;36:1011–1016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Durkin MJ, Dicks KV, Baker AW, Moehring RW, Chen LF, Sexton DJ, Lewis SS, Anderson DJ. Postoperative infection in spine surgery: does the month matter? J Neurosurg Spine. 2015;23:128–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gillespie WJ. The epidemiology of acute haematogenous osteomyelitis of childhood. Int J Epidemiol. 1985;14:600–606. [DOI] [PubMed] [Google Scholar]
- 9.Gruskay J, Smith J, Kepler CK, Radcliff K, Harrop J, Albert T, Vaccaro A. The seasonality of postoperative infection in spine surgery. J Neurosurg Spine. 2013;18:57–62. [DOI] [PubMed] [Google Scholar]
- 10.Haws BE, Braun BM, Creech TB, Barnard ER, Miller AN. Is there a seasonal influence on orthopaedic surgical wound infection rates? J Surg Orthop Adv. 2016;25:172–175. [PubMed] [Google Scholar]
- 11.Kane P, Chen C, Post Z, Radcliff K, Orozco F, Ong A. Seasonality of infection rates after total joint arthroplasty. Orthopedics. 2014;37:e182–e186. [DOI] [PubMed] [Google Scholar]
- 12.Klein EY, Sun L, Smith DL, Laxminarayan R. The changing epidemiology of methicillin-resistant Staphylococcus aureus in the United States: a national observational study. Am J Epidemiol. 2013;177:666–674. [DOI] [PubMed] [Google Scholar]
- 13.Leekha S, Diekema DJ, Perencevich EN. Seasonality of staphylococcal infections. Clin Microbiol Infect. 2012;18:927–933. [DOI] [PubMed] [Google Scholar]
- 14.McBride ME, Duncan WC, Knox JM. The environment and the microbial ecology of human skin. Appl Environ Microbiol. 1977;33:603–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mermel LA, Machan JT, Parenteau S. Seasonality of MRSA infections. PLoS One. 2011;6:e17925. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Namba RS, Inacio MCS, Paxton EW. Risk factors associated with deep surgical site infections after primary total knee arthroplasty: an analysis of 56,216 knees. J Bone Joint Surg Am. 2013;95:775–782. [DOI] [PubMed] [Google Scholar]
- 17.Perencevich EN, McGregor JC, Shardell M, Furuno JP, Harris AD, Morris JG, Fisman DN, Johnson JA. Summer peaks in the incidences of Gram-negative bacterial infection among hospitalized patients. Infect Control Hosp Epidemiol. 2008;29:1124–1131. [DOI] [PubMed] [Google Scholar]
- 18.Sagi HC, Cooper S, Donahue D, Marberry S, Steverson B. Seasonal variations in posttraumatic wound infections after open extremity fractures. J Trauma Acute Care Surg. 2015;79:1073–1078. [DOI] [PubMed] [Google Scholar]
- 19.Sahoo KC, Sahoo S, Marrone G, Pathak A, Lundborg CS, Tamhankar AJ. Climatic factors and community-associated methicillin-resistant Staphylococcus aureus skin and soft-tissue infections–a time-series analysis study. Int J Environ Res Public Health. 2014;11:8996–9007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Stott NS. Review article: Paediatric bone and joint infection. J Orthop Surg. 2001;9:83–90. [DOI] [PubMed] [Google Scholar]
- 21.Tanaka K, Kuwashima N, Ashizuka S-I, Yoshizawa J, Ohki T. Risk factors of infection of implanted device after the Nuss procedure. Pediatr Surg Int. 2012;28:873–876. [DOI] [PubMed] [Google Scholar]
- 22.Uçkay I, Betz M, Vaudaux P, Lauper N, Nicodème J-D, Abrassart S, Schindler M, Peter R, Christofilopoulos P. Is there a significant seasonality in the occurrence of osteoarticular infections? Infect Dis (Lond). 2015;47:252–254. [DOI] [PubMed] [Google Scholar]
- 23.Walling HW. Primary hyperhidrosis increases the risk of cutaneous infection: a case-control study of 387 patients. J Am Acad Dermatol. 2009;61:242–246. [DOI] [PubMed] [Google Scholar]