Electronic Letters to:

Hip: S. Kaspar and J. de V de Beer Infection in hip arthroplasty after previous injection of steroid J Bone Joint Surg Br 2005; 87-B: 454-457 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

Electronic letters published:

The safety of intra-articular injection of the hip joint

Birender Kapoor, Ahsaan Akhtar, Keith Barnes   (18 August 2005)

Dr Kaspar's reply to the letter from Zaki, Sadiq and Gambhir

Sam Kaspar   (1 August 2005)

Dr Kaspar's reply to letter from Atchia et al

Sam Kaspar   (1 August 2005)

Dr Kaspar's reply to letter from McGrory, Babikian and Crothers

Sam Kaspar   (1 August 2005)

Infection in total hip replacement after previous injection

Saeed H Zaki, Shahzad Sadiq, Anil K Gambhir   (21 July 2005)

Infection in hip arthroplasty after previous injection of steroid

Ismaël I Atchia, Mike Reed, Paul Partington, Derek Kramer and Fraser Birrell   (11 July 2005)

Steroid injection and infection after total hip arthroplasty

Brian J. McGrory, George Babikian, Omar D. Crothers   (1 July 2005)

The safety of intra-articular injection of the hip joint 18 August 2005
Birender Kapoor, Orthopaedic Higher Surgical Trainee Mersey Deanery, Ahsaan Akhtar, Keith Barnes Send letter to journal: Re: The safety of intra-articular injection of the hip joint beenu71{at}hotmail.com Birender Kapoor, et al.	Sir, We read this paper with interest. This is an interesting topic which requires further research in order to elucidate association of intra-articular injections with sepsis. However, we do not feel that this study supports the conclusions drawn by the authors. A ten fold increase in deep sepsis is highly significant and would make use of corticosteroid injections prior to arthroplasty a negligent practice! We believe that these results are partly the result of a sampling bias. It would also be interesting to note the exact time between injection and arthroplasty in the four proven cases of deep sepsis. We note that one of the patients from the 'infected and/ or infection tests' group underwent surgery two weeks following the injection. There is a paucity of literature on this subject, considering that intra-articular injections are so commonly performed. A study investigating the safety of corticosteroid injections in the hip joint noted a 6% rate of positive cultures from the tips of needles used with no clinical signs of sepsis. These injections were performed in operating rooms with full aseptic precautions. From the discussion of the article it appears that you believe a breach in asepsis to be the cause of infection whereas the abstract states this to be immunosuppression. The former is more likely as these injections were performed outside operating room conditions with some of the patients undergoing arthroplasty within weeks of having intra-articular injections. We hope it does not lead the readers to believe that an intra-articular injection at a reasonably distant time prior to arthroplasty is a negligent practice. B. KAPOOR A. AKHTAR K. BARNES University of Liverpool, Liverpool, UK.
Dr Kaspar's reply to the letter from Zaki, Sadiq and Gambhir 1 August 2005
Sam Kaspar, MD, Orthopaedic Surgeon McFarland Clinic, Ames Iowa USA Send letter to journal: Re: Dr Kaspar's reply to the letter from Zaki, Sadiq and Gambhir mightysamster1{at}aol.com Sam Kaspar	Sir, There were no deep hip infections in the non-steroid group. The figure 7.5% included superficial wound infections and patients who had poor results, but who did not indicate findings of deep infection. Regarding asepsis at hip surgery, we used modern facilities with all the precautions listed. The overall rate of revision for all causes was 1.02% in the database, with Kaplan-Meier survivorship extrapolated out to over four years. To put this into perspective, the four deep hip infections and one revision for dislocation in the group of only 40 steroid-arthroplasty patients, would compare most unfavourably with the ten revisions (most of which were not for infections) in about 1000 non-steroid patients. I would distinguish between overt infections after the injection and late infection after subsequent arthroplasty. Low-grade hip infections may not always be evident immediately as they would in a knee, which is more superficial and prone to visible effusion. For a long time steroids have been known to cause problems in hips with osteonecrosis, and I believe that the hip joint differs from most other joints in terms of its response to steroid loads. S. KASPAR, MD McFarland Clinic, Iowa, USA.
Dr Kaspar's reply to letter from Atchia et al 1 August 2005
Sam Kaspar, MD, Orthopaedic Surgeon McFarland Clinic, Ames Iowa USA Send letter to journal: Re: Dr Kaspar's reply to letter from Atchia et al mightysamster1{at}aol.com Sam Kaspar	Sir, I thank the correspondents for their interest and their comments. They concur with our interpretation concerning the potential for patient-selection bias, the uncertainty about which part of the injection procedure may have led to the striking results, and the statistical limitations of a relatively small, single-centre retrospective study. They re-organise our data in a different way which distorts the statistical outcome, but we are aware of the limitations of statistics as a tool, particularly when attempted re-calculations are fraught with significant beta error (underpowered due to small sample size). Infections in hip arthroplasty are so rare that one would require a large study in order to demonstrate a difference in small rates of infection. Statistical shuffling cannot disprove the rate of infection in the steroid group is unacceptable. I (and major journals dealing with arthroplasty outcomes) see merit in the survivorship curve method, and I think the figure is self-explanatory in terms of its gravity regarding the concerns over the injection procedure. Our study demonstrated sufficient concern for an audit-style retrospective review to be undertaken before advocating the procedure, or any prospective evaluation of it. I am unsure as to why the correspondents underestimate the dramatic rate of infection in the steroid group, yet they note that 'there is an ever increasing need to investigate the safety and efficacy of intra-articular injections in patients with osteoarthritis.' The literature would not support that assertion pertaining to therapeutic efficacy of steroid hip injections. The 'potential' for this therapy, with the insinuation that it should be used more to alleviate the increasing need for arthroplasty, is a question which was asked over five decades ago (and largely answered with a 'no' in subsequent studies). The existing published studies on the subject of efficacy will be the subject of a forthcoming manuscript. I removed, for brevity, extensive material from the infection article regarding acceleration of arthritic decay and lack of efficacy of the injection procedure. I will probably prepare a separate manuscript on those subjects. In summary, while I commend the correspondents for their interest in the statistics, I do not feel that their calculations add anything to the debate. Also, I think we need to focus on clinical significance and not digress into statistical evaluations of data. Finally, multiple centres should review their results retrospectively and report them. I would also recommend that the correspondents review the published data on the lack of efficacy of steroid hip injections, prior to asserting that they are therapeutically promising; that is merely an unfounded opinion. For the reasons outlined above, however, I do not agree with their comments. S. KASPAR, MD McFarland Clinic, Iowa, USA.
Dr Kaspar's reply to letter from McGrory, Babikian and Crothers 1 August 2005
Sam Kaspar, MD, Orthopaedic Surgeon McFarland Clinic, Ames Iowa USA Send letter to journal: Re: Dr Kaspar's reply to letter from McGrory, Babikian and Crothers mightysamster1{at}aol.com Sam Kaspar	Sir, We thank McGrory, Babikian and Crothers for their comments and the evaluation and summary of their own series. I look forward to reading more reports of a similar nature in order to gain some idea as to whether a multi-centre trend is present. S. KASPAR, MD McFarland Clinic, Iowa, USA.
Infection in total hip replacement after previous injection 21 July 2005
Saeed H Zaki Wrightington Hospital, Wigan, Shahzad Sadiq, Anil K Gambhir Send letter to journal: Re: Infection in total hip replacement after previous injection saeedzaki{at}hotmail.com Saeed H Zaki, et al.	Sir, This paper highlights the significantly increased rate of revision for infection (12.5%) after total hip replacement (THR) in patients who had undergone an intra-articular injection of steroid. The unusually high (7.5%) rate of infection in patients who did not receive an injection is of even more concern. The risk of infection after THR is less than 1%, if surgery is performed in a laminar flow theatre, using prophylactic antibiotics and an exhaust suit.1 Although the authors have mentioned the aseptic protocol for hip injections they do not give details of their protocol for total joint replacements. Do they use laminar airflow with exhaust suits? We would also like to know if the authors can think of any mitigating circumstances that could have contributed to this higher than usual rate of deep infection? In a prospective study,2 40 patients underwent hip aspiration and injection by Orthopaedic trainees and consultants using strict aseptic protocol. None of the patients developed deep infection and contamination was reported in only two. Do the authors feel that the results of their study can be extrapolated to other centres and if so, should intra-articular injections be abandoned as a procedure? Can the results of their study be applied to other joints, particularly knees, which are injected much more often and in much less sterile an environment than are hips? In a survey regarding the rate of infection after steroid injections of the knee joint, the risk of infection was found to be very low.3 We feel that hip injections are a useful diagnostic and therapeutic tool and with a strict aseptic technique the infection rate is no higher in hip replacements with injection than in those without injection. S. H. ZAKI S. SADIQ A. K. GAMBHIR Wrightington Hospital, Wigan, UK. 2. Lidwell OM, Lowbury EJL, Whyte W, et al. Effect of ultra clean air in operating rooms on deep sepsis in the joint after total hip or knee replacement: a randomised study. Br Med J 1982;285(6334):10-14. 2. Waseem M, Sadiq S, Gambhir AK, et al. Safety and efficacy of intra-articular injection of the hip. Hip International 2002;12:378-82. 3. Charalambous CP, Tryfonidis M, Sadiq S, Hirst P, Paul A. Septic arthritis following intra-articular steroid injection of the knee: a survey of current practice regarding antiseptic technique used during intra-articular steroid injection of the knee. Clin Rheumatol 2003;22:386-90.
Infection in hip arthroplasty after previous injection of steroid 11 July 2005
Ismaël I Atchia, Clinical Research Fellow Northumbria Healthcare NHS Trust, Mike Reed, Paul Partington, Derek Kramer and Fraser Birrell Send letter to journal: Re: Infection in hip arthroplasty after previous injection of steroid ish{at}doctors.org.uk Ismaël I Atchia, et al.	Sir, We read this article with interest. Kaspar and de V de Beer raise concerns about the safety of injecting steroids into patients with osteoarthritis who may be candidates for total hip arthroplasty. They conclude that in these patients intra-articular injection of steroid is contra-indicated. However, there are methodological problems which limit the credibility of this conclusion. This was a single centre retrospective audit, and as the authors themselves state, a multi-centre study would be required to corroborate the findings. Studies based on retrospective identification of cases are fraught with problems of interpretation, particularly selection bias. The injected cohort may have been selected for severity or duration of disease prior to arthroplasty, leading to a technically difficult procedure and a worse outcome.1,2 Therefore, the injection group, regardless of having an injection, could already be at a higher risk of post–operative complication and the study findings a consequence of this existing increased risk. Given that the injections in the study were undertaken by multiple radiologists under fluoroscopy, aspects of individual technique or the use of contrast may have been associated with a risk of infection rather than the steroid. There was also an error of classification in that one of five ‘revision’ cases did not require implant removal and a pathogen was not identified. Therefore the appropriate comparison would be four revisions (injection group) vs one revision (comparison group). Using a two-tailed Fisher exact test (for small group comparison without a directional prior hypothesis) the p value (> 0.35) does not approach statistical significance. Similarly, the comparison of confirmed infections of three (injection group) vs nil (comparison group) has a p > 0.24 (two-tailed Fisher exact). This confirms that the study was not designed to demonstrate statistically or clinically significant differences in rates of infection or revision. This re-analysis of the key findings may help to interpret the apparently poor face validity of their hypothesis. The authors, based on their description of the cases, imply that injection between two weeks and 29 months (assuming that the missing units in Table II are months) prior to injection can result in infection up to three years after surgery (one of the cases had infection after 35.7 months). It seems surprising to have such late sequelae from an injection. More definitive data is needed before we discard such a promising therapeutic option, particularly as ultrasound guidance now offers accuracy without the risks of ionising radiation or contrast.3,4 As the prevalence of osteoarthritis and the requirement for total hip arthroplasty increases,5 there is an ever increasing need to investigate the safety and efficacy of intra-articular injections in patients with osteoarthritis. While provocative, the study by Kaspar and de V de Beer has significant limitations, and further studies are required to answer the questions raised. I. I. ATCHIA M. REED P. PARTINGTON D. KRAMER F. BIRRELL Northumbria Healthcare NHS Trust, Northumberland, UK. 1. Wymenga A, van Horn JR, Theeuwes A, Muytjens HL, Slooff TJJH. Peri-operative factors associated with septic arthritis after arthroplasty. Acta Orthopaedica Scandinavica 1992;63:665-71. 2. Kaltsas D. Infection after total hip arthroplasty. Annals R Coll Surg Engl 2004;86:267-71. 3. Qvistgaard E, Kristoffersen L, Terslev et al. Guidance by ultrasound of intra-articular injections in the knee and hip joints. Osteoarthritis and cartilage 2001;9:512-17. 4. Karim Z, Brown AK, Quinn M, et al. Ultrasound-guided steroid injections in the treatment of hip osteoarthritis: comment on the letter by Margules. Arthritis Rheum 2004;50:338-9; author reply 339-40. 5. Birrell F, Johnell O, Silman AJ. Projecting the need for hip replacement over the next three decades: influence of changing demography and threshold for surgery. Ann Rheum Dis 1999;58:569-72.
Steroid injection and infection after total hip arthroplasty 1 July 2005
Brian J. McGrory, Orthopaedic Surgeon Orthopaedic Associates of Portland, Maine, USA, George Babikian, Omar D. Crothers Send letter to journal: Re: Steroid injection and infection after total hip arthroplasty mjri{at}yahoo.com Brian J. McGrory, et al.	Sir, We read this paper with both interest and concern. It would have been helpful if the authors had clarified the time interval between injection and surgery in Table II; the units are not labelled in either the table or in the text. The alarming finding by the authors prompted us to assess the deep infection rate in 82 consecutive patients who underwent injection prior to ipsilateral total hip replacement in our practice. The technique and content of hip injection is similar to that in the authors’ institution, with surgeons performing injections under sterile conditions with fluoroscopic guidance. One exception was that one of the physicians in our practice did not routinely use radio opaque contrast medium prior to injection of steroid/anaesthetic medication (8/95 injections, no infections). In 82 patients who had one or more injections (two had three prior injections, nine had two and 71 had one prior injection) between 1995 and present, there were two deep infections. One infection caused by staphylococcus epidermidis was indolent, and was diagnosed at 17 months post-operatively; the other was an acute staphylococcus aureus infection diagnosed at three years after operation. In both cases, the injection had been administered four months before the index surgery. We compared this rate of infection (2/82) with a random sample of 211 patients from a previously performed outcome study in our office (1995-2001). There were no deep infections in this cohort. Statistical analysis using the Fisher exact test demonstrated no significantly higher rate of infection in the injection group (p = 0.078). Despite the lack of statistical significance, the paper by Kaspar and de V de Beer has raised our awareness of the possibility of an association between total hip arthroplasty infection and pre-operative steroid injection. While further study in both the hip and knee replacement is warranted to confirm the authors' conclusions, we thank them for bringing this finding to our attention. B. J. McGRORY, MD G. BABIKIAN, MD O. D. CROTHERS, MD 33 Sewall Street, Portland, Maine 04102-1260, USA