Positive Intraoperative Cultures and Canine Total Hip Replacement: Risk Factors, Periprosthetic Infection, and Surgical Success

Introduction

Clinically significant microbial damage of the tissues surrounding the hip prosthesis, referred to in this paper simply as “infection,” is a serious complication of canine total hip replacement (THR).¹² It results in fair to poor hip function, and its resolution requires permanent removal of the hip prosthesis.³ Reported predisposing factors for infection in dogs include surgical revision of THR, bacteremia secondary to a localized infection in a remote area, hip infection prior to surgery, Brucella canis infection, and prolonged surgery time.^3–6 Additional predisposing factors in the human literature include air and contact contamination, traumatic surgical technique, and instability.¹⁷⁸ Preoperative case assessment and meticulous attention to asepsis and surgical technique can help to eliminate or minimize these factors. The incidence of infection can also be reduced by early identification and treatment of surgical site contamination (i.e., contamination being the presence of bacteria within a wound without bacteria-associated tissue damage). For many years, in both humans and dogs, intraoperative cultures of the hip joint taken just prior to closure have been used to identify surgical site contamination. In humans, the usefulness of these cultures is controversial. Schneider reported that they were of no value; he believed that the presence of an uncontaminated surgical wound was practically nonexistent and found that approximately 15% of cases with sure clinical signs of infection had negative intraoperative cultures.¹ Conversely, Lidwell, et al., found that there was a significant association between the isolation of Staphylococcus aureus (S. aureus) from intraoperative cultures and subsequent infection; they found that 8% of cases, from which S. aureus was isolated, developed infection.⁷ To the authors’ knowledge, no study has been conducted on the usefulness of intraoperative cultures taken during canine THR.

This retrospective study reviewed 27 canine THR operations. The purpose was to assess the relationship between the results of closing intraoperative cultures and presurgical and surgical factors, and short-and long-term outcome.

Materials and Methods

Thirty-five THR operations using the canine cemented modular hip system^a were performed in 33 dogs between January 1994 and May 1997. A study population of 27 dogs was selected from these cases to include only those dogs for which bilateral hip dysplasia with secondary osteoarthritis was the sole indication for THR, and dogs that did not have hip surgery prior to THR. These inclusion criteria resulted in a homogenous case population with respect to the preoperative state of the operated hip.

All of the cases were screened prior to surgery for localized and systemic infections. This screen included a complete clinical history, a physical examination, routine hematological and serum biochemical evaluation, and hip radiography. The surgical procedure used has been previously described.⁹ In all cases, aseptic technique was meticulous, and extra precautions were taken to avoid surgical site contamination relative to other orthopedic operations carried out at the same hospital. Total hip replacements were performed at the beginning of the day in an operating room with forced planum ventilation that was reserved for ultraclean orthopedic operations. After the first incision, no persons were permitted to enter the operating room. Waterproof gowns and drapes were used, and the surgical site was draped with an adhesive, iodine-impregnated, plastic drape.^b

All animals received perioperative intravenous (IV) antibiotics, and 1 g of cefazolin sodium^c was added to each 20 g of polymethylmethacrylate cement. Prior to closure, the surgical site was flushed with sterile saline, and then two swabs of the hip joint were taken for aerobic and anaerobic culture. Sterile saline flushing was a routine practice in surgery, used to reduce bacterial contamination and to remove blood clots and debris.⁸ Antibiotic sensitivity testing was carried out for all aerobic isolates using 12 or 13 antibiotics. An experienced microbiologist assessed the sensitivity pattern for each isolate as being normal for the species of bacteria cultured, as demonstrating increased resistance, or as demonstrating multidrug resistance. All cases with positive intraoperative culture results were treated with oral antibiotics for at least 2 weeks on the basis of culture and sensitivity results. Fifty percent of cases with negative cultures were treated with a maximum of 5 days of antibiotics postoperatively, pending final intraoperative culture results. The remaining cases with negative culture results were not treated with oral postoperative antibiotics, as preliminary culture results strongly suggestive of the absence of bacteria were available for these cases within 24 to 48 hours of surgery. Perioperative analgesia was provided as necessary for all cases.

Radiographs of the hip prosthesis were taken immediately after surgery for all cases and at short-term follow-up for all those cases that were included in the analysis of short-term success. A board-certified radiologist read these radiographs, and any signs consistent with infection were noted. These signs have been previously described.¹⁰¹¹ Intraoperative and postoperative complications of THR were recorded.

The 27 dogs were divided into two groups: those with positive intraoperative culture results and those with negative results. These two groups were then compared in three stages to analyze the relationship between intraoperative culture results and presurgical, surgical, and postsurgical factors. All the statistical tests described were carried out using SPSS 10 software.^d A value of P<0.05 was considered significant.

The following presurgical and surgical variables were obtained from the case records: age, body weight, duration of the clinical signs of hip disease prior to THR, size of the prosthetic acetabulum and femoral stem, duration of surgery, and class of perioperative antibiotic used. The continuous variables were assessed for normality using the Shapiro-Wilk test. The t-test or Mann-Whitney U test was used to determine whether these variables differed significantly between those cases with positive culture results and those with negative results.

Information regarding short-term outcome of THR was then reviewed. As concurrent orthopedic and neurological disease can affect the perceived success of THR, cases with such diseases were excluded from the assessment of short-term success. Short term was defined as the time between THR and the first follow-up examination between the 4th and 16th weeks postoperatively. For each case, grade of lameness at the time of discharge, grade of lameness and activity at follow-up examination, and the degree of improvement in lameness between discharge and follow-up examination were recorded. The grading systems for these factors are given in Tables 1 and 2. The Mann-Whitney U test was used to determine whether these factors differed significantly between those cases with positive culture results and those with negative results. In addition, the times of discharge and short-term follow-up relative to the operation date were recorded. These times were tested for normality using the Shapiro-Wilk test and were compared between positive and negative culture groups using the t-test or Mann-Whitney U test as appropriate. The latter was conducted to determine whether these times might have confounded the analysis of short-term success.

Finally, information regarding long-term outcome following THR was collected. Reoperation following THR increases the risk of periprosthetic infection. Therefore, cases reoperated in the short term were excluded for the assessment of long-term success. Cases were included if long-term follow-up of at least 1 year was available in the medical records or from telephone conversations with owners. A two-way contingency table analysis was conducted to evaluate whether the proportions of cases with long-term problems differed between those cases with positive culture results and those with negative results.

Results

Data from the 27 dogs that met the study’s criteria is summarized in Tables 3 and 4. Their mean age was 57±40 months, and their mean weight was 40.1±9.9 kg. The mean duration of the clinical signs of hip disease prior to THR was 13.0±12.8 months. None of these cases showed signs of active infection preoperatively, and the results of routine biochemical and hematology tests were within the reference ranges. Cefazolin sodium,^c cefoxitin sodium,^e and cephalexin^f were the perioperative antibiotics used in 15, seven, and five of the cases, respectively. Regardless of which antibiotic was used, all cases received a minimum of 20 mg/kg IV doses at induction and every 2 hours during surgery, except case no. 3 which received 10 mg/kg doses of cefoxitin sodium. Where cefoxitin sodium was used, an equal intramuscular dose was given at induction. The mean duration of surgery was 134±21 minutes.

Eleven out of 27 cases had positive intraoperative culture results [Table 5]. Age, body weight, duration of surgery, and class of perioperative antibiotic used did not differ significantly between those cases with positive culture results and those with negative results. The mean duration of the clinical signs of hip disease prior to THR was significantly greater (P=0.034) for those cases with positive culture results (mean, 20.3±16.3 months) than for those cases with negative culture results (mean, 8.0±6.3 months).

Eighteen cases (eight with positive culture results and 10 with negative results) were included in the statistical analysis of short-term success. The mean discharge time for these cases was 4.3±1.8 days postsurgery, and the mean time of follow-up examination was 8.2±3.3 weeks. These times did not differ significantly between cases with positive culture results and cases with negative results. The median grade of lameness at discharge and follow-up was 3 and 1, respectively. Therefore, lameness improved by a median of two grades over time. The mean grade of activity at follow-up was 2. There were no significant differences between those cases with positive intraoperative culture results and those cases with negative results with respect to lameness at discharge and follow-up, activity at follow-up, and improvement in lameness between discharge and follow-up.

No intraoperative complications were recorded. The only short-term complication recorded was luxation of the prosthesis, which occurred in one case (case no. 19), and this was not associated with a positive intraoperative culture result. Despite luxation of the prosthesis at short-term follow-up, this case favored the unoperated limb. This was due to severe subluxation of the unoperated hip. Surgical reduction was associated with a negative intraoperative culture result. Five months later, a diagnosis of periprosthetic infection with a coagulase-negative Staphylococcus spp. was made. This resulted in removal of the prosthesis. (This case was excluded from the following long-term success analysis.)

Fifteen cases (seven with positive culture results and eight with negative results) were included in the statistical analysis of long-term success. Mean follow-up time was 25.5±3.0 months. Of those cases with positive culture results, six (85.7%) had no prosthesis-associated problems, and one (14.3%) had periodic stiffness and lameness on the prosthetic limb, for which further diagnostics were declined by the owner. Of those cases with negative culture results, five (62.5%) had no prosthesis-associated problems, two (25.0%) had a persistent lameness on the prosthetic limb, and one (12.5%) had aseptic loosening of the femoral stem. The latter resulted in surgical removal of the prosthesis, which was associated with a negative culture result and no evidence of postoperative infection. Again, further diagnostics for the two cases with persistent lameness were declined because of the mildness of the signs. There was no significant difference in the incidence of long-term problems between those cases with positive culture results and those cases with negative results.

Infection was not diagnosed in any of the 27 cases assessed, apart from the aforementioned case following surgical reduction of a luxated hip prosthesis.

Discussion

This retrospective study analyzed the results of closing intraoperative cultures from 27 canine THR operations performed for the management of bilateral hip dysplasia. Eleven (40.7%) of these cultures were positive. A wide variety of organisms were isolated. This variation has been noted in previous studies of THR.¹⁸¹² The organisms ranged from being known pathogens (e.g., Clostridium spp.) to nonpathogens (e.g., Acinetobacter lwoffi).¹³ However, none of these organisms can be dismissed as being harmless in the immunocompromised environment of a perioperative total hip prosthesis.

The high positive culture rate is concerning because of the possible subsequent high rate of infection. In the cases studied, a number of measures were taken to minimize surgical site contamination and periprosthetic infection, as outlined in the Materials and Methods section. However, surgery time was ≥90 minutes in all cases, and this factor most likely contributed to the high positive culture rate.¹⁴

The advantages of using both prophylactic perioperative IV antibiotics and antibiotic-impregnated cement in THRs in dogs and humans are well recognized.⁸¹² For approximately half of the cases in this study, cefazolin sodium was administered at 20 mg/kg body weight, IV, every 2 hours during surgery. This antibiotic regimen has been shown to produce adequate prophylactic concentrations in the tissues exposed during THR.¹⁵ Intravenous cefoxitin sodium or cephalexin was used in the remainder of the cases. The use of these antibiotics did not significantly affect the culture results, but the numbers of cases receiving each of these antibiotics were small. Therefore, definite conclusions cannot be made regarding their efficacy. In humans, gentamicin is the most popular antibiotic used in the cement. One of the main reasons for this popularity is its powerful effect against S. aureus.¹² Staphylococcus spp. are the most commonly cultured bacteria from surgical wounds in dogs and cats.¹⁶¹⁷ Therefore, gentamicin may also be the preferred antibiotic for use in the cement for canine THR. Further studies are required to compare gentamicin and cefazolin sodium in this setting.

The mean duration of the clinical signs of hip disease prior to THR was significantly greater (P=0.034) for those cases with positive culture results than for those cases with negative culture results. This relationship may have been found to be significant simply by chance and may have no underlying meaning; however, a possible explanation for this relationship is as follows. If the duration of the clinical signs in dysplastic dogs is positively related to the severity of the osteoarthritis, then the severity of the osteoarthritis may adversely alter the pharmacokinetics of perioperative antibiotics at the surgical site and thus allow the establishment of contaminating organisms. Conversely, Marcellin-Little showed that there was no significant difference in the distribution of IV cefazolin sodium in the hip joint between dogs with and without osteoarthritis.¹⁵

None of the cases with positive intraoperative cultures were diagnosed with periprosthetic infection. Surgical site contamination, as identified by a positive culture, may not have resulted in clinically significant infection for the following reasons: the contamination responded to a 2-week course of oral antibiotics, the detailed follow-up was too short,¹⁸ too few cases were included in the study, subclinical infection was present, the contaminants were not of sufficient pathogenicity, or a single intraoperative culture taken from the hip joint just prior to closure was not an accurate predictor of infection. The latter has been documented in humans, and so a number of ways to improve the predictive value of intraoperative cultures have been considered. It has been reported that a single positive intraoperative culture of the hip joint taken either immediately after the joint capsule was opened (i.e., an opening culture) or just prior to closure (i.e., a closing culture) was not a significant predictor of subsequent infection, but that the combination of a positive opening culture and a positive closing culture was a significant predictor of subsequent infection.¹⁹ Other investigators have found that the number and species of contaminating organisms are important. Specifically, it has been reported that patients with bacterial counts in the top 30% had a nearly twofold increased risk of developing subsequent infection relative to other patients, and that the isolation of S. aureus was associated with a seven-fold increased risk of infection.⁷

It has been suggested that the incidence of suspected periprosthetic infections in humans may be equal to that of diagnosed infections.⁷ A suspicion of infection was based on the presence of mild abnormal pain, which was not severe enough to justify reoperation. These investigators found that the incidence of suspected infections and confirmed infections decreased in parallel with the introduction of perioperative antibiotics and an ultraclean operating environment.⁷ In dogs, abnormal pain may present as persistent lameness or decreased activity. In this study, these two factors were not found to be significantly related to intraoperative culture results. In addition, despite the high incidence of positive cultures, the percentage of cases with clinically excellent function in the long term (73.3%) was comparable to that of previous studies.³²⁰

In dogs, other ways of identifying surgical site contamination apart from intraoperative cultures are not widely used. In humans, percutaneous needle aspiration has been reported to be an accurate means of identifying bacterial contamination.²¹ However, this procedure is associated with the risk of introducing bacteria into a sterile joint. In this study, subclinical infection may have been present among those cases with positive intraoperative cultures, despite a 2-week course of antibiotics. Some surgeons recommend a 6- to 8-week course of antibiotics for cases with positive intraoperative culture results.⁹

Conclusion

The results of the closing intraoperative culture results from 27 canine THR operations were assessed. All cases received intraoperative antibiotics and postoperative antibiotics, pending culture results, and cases with positive cultures received a minimum of 2 weeks of postoperative antibiotics based on the results of sensitivity testing. Eleven cases had positive culture results, but periprosthetic infection was not diagnosed in any of these cases. The mean duration of the clinical signs of hip disease prior to THR was significantly greater for those cases with positive cultures (P=0.034). Culture results were not significantly related to surgical success in all 27 dogs studied. Other studies conducted of human periprosthetic infections assessed approximately 2,000 and 8,000 THR patients, respectively, to obtain statistically significant results.¹⁷ Such large-scale studies are ideal but expensive. Further studies including a larger number of cases are needed to confirm these findings. In addition, the incidence of subclinical infection and its effects on short- and long-term surgical success need to be addressed.

The findings of this study suggest that the results of closing intraoperative cultures do not have implications for the success of THR beyond the immediate postsurgical period, if strict measures are employed to minimize the development of periprosthetic infection and 2-week courses of postoperative antibiotics are administered to those cases with positive cultures. This paper, therefore, confirms the benefits of the following: careful preoperative case assessment for preexisting infections; meticulous attention to asepsis in the operating room; antibiotic-impregnated cement; the use of perioperative antibiotics that have been shown to penetrate the THR surgical site (e.g., cefazolin sodium); closing intraoperative cultures; and a 2-week postoperative course of antibiotics in cases of positive closing cultures. In addition, the owners of dogs presenting for THR following a prolonged period of clinical signs (i.e., >20 months) should be advised that the risk of periprosthetic infection may be greater.