Sterile Versus Nonsterile Gloves: Infection Risk and Cost Comparison in Corticosteroid Shoulder Injections

Abstract

Background:

Corticosteroid injection into the shoulder is a common treatment used for various orthopaedic pathologies. However, it remains unclear if using nonsterile gloves correlate with higher infection risk compared with sterile gloves for such procedures. Although literature in other fields supports the noninferiority of nonsterile gloves toward infection risk, similar studies have not been conducted in orthopaedics. This study seeks to identify if there is an increased infection risk and difference in costs with using nonsterile gloves compared with sterile gloves for shoulder injections.

Methods:

A retrospective review of medical records identified patients who received shoulder injections from a single surgeon from February 2023 to July 2024. The surgeon transitioned from using sterile gloves to nonsterile gloves during this period, and a proportionate number of patients were assigned to nonsterile and sterile glove groups. Post hoc power analysis was completed through calculating confidence intervals and minimal detectable effect size through Cohen h. Records were reviewed for demographic information, body mass index, Charlson comorbidity index, follow-up length, and postinjection infection as identified through ICD-10 code T80.29XA. Rate of infection was evaluated at 1, 3, and 6 months following injections. Glove prices were obtained through our institution. Percentages were obtained for categorical variables, and means were calculated for continuous variables. Descriptive statistics were done using Chi-squared tests for percentages and Student t tests for means for any baseline differences between the patient groups.

Results:

A total of 641 patients were evaluated with a minimal follow-up period of 6 months. Three hundred seventeen injections were performed using sterile gloves, and 323 injections were performed using nonsterile gloves. No shoulder joint infections were recorded following any injection performed with either glove type, with the study being sufficiently powered to detect within 0.012 infection rate difference. Sterile gloves were 36 times more expensive than nonsterile gloves per injection.

Conclusion:

Although sterile gloves are substantially more expensive than nonsterile gloves, no difference was found in infection risk outcomes for corticosteroid shoulder injections in clinical practice.

Local corticosteroid injections are commonly used for managing musculoskeletal pathologies in orthopaedic practice. Conditions, such as osteoarthritis, bursitis, and tendinopathy, are often managed with cortisone injections as an adjunct to physical therapy and anti-inflammatory medication.^1,2 For shoulder pathology, corticosteroid is injected into the bursal space above the rotator cuff and below the acromion for the treatment of bursitis, rotator cuff tendinitis, and anterior superior impingement syndrome of the shoulder.^3,4 It can also be injected directly into the glenohumeral joint. However, this technique carries the risk of infection if not properly performed. The seeding of a pathogenic organism in a joint resulting in septic arthritis presents a rare but potentially disastrous complication, making instrument sterility an important concern.⁵

There is no consensus among providers regarding the use of sterile versus nonsterile gloves for corticosteroid shoulder injections. Existing orthopaedic literature is unclear and lacks direct clinical comparisons.^6,7 Although limited, more definitive findings appear among other medical specialties. In Mohs surgery, infection rates have been shown as similar between nonsterile and sterile gloves.^8-11 Comparable results have been demonstrated with nonsterile dental extractions.^12-14 Finally, multiple studies of minor outpatient procedures, such as skin biopsies and uncomplicated laceration repairs, have provided evidence for nonsterile gloves carrying no greater infection risk than sterile gloves.^15-18 Given that sterile gloves can incur approximately double the price in resources than nonsterile gloves, confirming the relative safety of nonsterile gloves would provide a notable benefit to both patient and clinical practice.¹⁹

This study intends to extend the findings in the literature to the field of orthopaedics. Specifically, it will seek to compare nonsterile gloves to sterile gloves for infection outcomes in corticosteroid shoulder injections while also evaluating cost utilization. In doing so, the results obtained will help guide clinical practice toward the best balance of safety standards and cost-efficiency.

Methods

Following institutional review board approval, a retrospective review of medical records identified patients who received shoulder injections from a single fellowship-trained shoulder and elbow surgeon from February 2023 to July 2024. These patients were treated during the period in which the surgeon transitioned from using sterile gloves and technique to nonsterile gloves for injections. No changes occurred in the surgeon's area of practice, institutional infection policies, or patient population during this time. The surgeon's standard technique also remained unchanged, involving the preparation of the site using chlorhexidine gluconate with alcohol and the injection of the subacromial space from a lateral approach. A proportionate number of patients was assigned to the sterile gloves and nonsterile gloves groups. The power of the selected sample size was evaluated using post hoc analysis. This involved the calculation of confidence intervals for each group's rate and minimal detectable effect size through Cohen h.²⁰ Electronic health records were reviewed for demographic information, body mass index (BMI), Charlson comorbidity index (CCI), follow-up length, and postinjection infection as identified through International Coding of Diseases code T80.29XA. Rate of infection was evaluated at 1, 3, and 6 months following injections. In addition, the medical records of all involved patients were independently reviewed for any clinically identified infections. A representative of our institution was contacted to obtain specific information on the prices of sterile and nonsterile gloves. Percentages were obtained for categorical variables, and averages were calculated for continuous variables. Descriptive statistics were done using Chi-squared tests for percentages and Student t tests for means of parametric data.

Results

Demographics

A total of 641 patients were evaluated with a minimum follow-up of 6 months. Three hundred seventeen injections were performed using sterile gloves and 323 injections were performed using nonsterile gloves; 59.4% of the patients were female and had an average age of 64.8 years, average BMI of 30.5 kg/m², and average CCI of 4.5. In addition, 85.5% of the patients were Caucasian/White. No notable differences were observed in sex, race, average BMI, or average CCI between the two groups. Further details on demographics can be found in Table 1.

Table 1.

Comparison of Demographic Characteristics Between Patients Receiving Shoulder Injections With Sterile and Nonsterile Gloves

Characteristics	Sterile Gloves (n = 317)	Non-Sterile Gloves (n = 323)	P Value (95% CI)
Mean age	65 yr	66 yr	0.521 (−1.39-2.74)
Sex
Female	192 (61%)	196 (61%)	0.990
Male	125 (39%)	127 (39%)	0.990
Mean BMI	30.2 kg/m2	30.8 kg/m2	0.286 (−58.34-3.98)
Mean CCI	4.6	4.3	0.347 (−0.28-0.81)
Race
Asian	7 (2%)	7 (2%)	0.964
African-American/Black	58 (18%)	64 (20%)	0.676
Hispanic	55 (17%)	54 (17%)	0.834
Caucasian/White	180 (57%)	185 (57%)	0.906
Other	17 (5%)	13 (4%)	0.487

BMI = body mass index, CCI = Charlson comorbidity index, CI = confidence interval

Postprocedure Infection Rate

Following corticosteroid shoulder injections, no infections were recorded at any time point within the 6 months of follow-up in both the sterile and nonsterile glove groups. Using the rule of three for zero events, the confidence intervals for infectious rates in both groups were less than 1%. Given the chosen sample sizes, the smallest detectable difference in infection rate was 0.012 with an assumed power level of 0.80. Further information on this analysis can be found in Table 2.

Table 2.

Comparison of Infection Rates Between Patients Receiving Shoulder Injections With Sterile and Nonsterile Gloves Alongside Results of Power Analysis

	Sterile Gloves (n = 317)	Nonsterile Gloves (n = 323)
Infection rate	0.00	0.00
CI	−0.95-0.95	−0.93-0.93
Cohen h	0.22
Minimal detectable difference	0.012	0.012

CI = confidence interval

Cost Comparison

At our institution, the cost per box of 50 sterile gloves is $53.93 and the cost per box of 200 Nitrile nonsterile gloves is $5.75. Assuming only one pair of gloves was used per subacromial corticosteroid injection procedure, a total of $683.83 was spent for sterile injections, which equates to $2.16 per procedure. In comparison, a total of $18.57 was spent for nonsterile injections, which equates to $0.06 per procedure. A comparison of cost utilization between the two glove groups is outlined in Table 3.

Table 3.

Comparison of Cost Between Shoulder Injection Procedures Using Sterile and Nonsterile Gloves

	Sterile Gloves (n = 317)	Nonsterile Gloves (n = 323)
Price per box	$53.93	$5.75
Price per glove	$1.07	$0.03
Total procedural cost	$683.83	$18.57
Cost per procedure	$2.16	$0.06

Discussion

The purpose of this study was to compare the infection outcomes and cost utilization in patients who underwent corticosteroid shoulder injections with sterile gloves and nonsterile gloves. Demographic and comorbidity risk factors were similar between patients from both groups. No infections were reported in this study population, although we found a substantial cost burden when using sterile gloves relative to nonsterile gloves. The selected sample size was also sufficiently powered to detect any clinically relevant differences.

The findings are in line with the literature from other medical specialties, which heavily support the noninferiority of nonsterile gloves in infection risk compared with sterile gloves.^8-14 Some preclinical studies have identified the propensity for bacteria to colonize nonsterile gloves, especially those left in open glove boxes exposed to air.²¹ However, in clinical practice, this proposed bacterial load has not been shown to reach levels high enough for clinically meaningful infection.¹⁷ In addition, research for minor surgical procedures, such as laceration repair and wound débridement, has further supported the application of these findings. Two recently published systematic reviews and meta-analyses of randomized clinical trials across outpatient, primary care, and emergency department settings each found no statistically significant difference in infection rate with either glove choice.^22,23

Beyond infection risk, the shift toward nonsterile gloves may also carry economic and environmental implications. Financially, additional retail costs of materials for sterile gloves drive a substantial difference to nonsterile gloves. At our institution, the price of sterile gloves is 838% more expensive than nonsterile gloves per box. This equates to a 3500% cost increase when performing a subacromial corticosteroid injection with a pair of sterile gloves as nonsterile gloves. If all of the procedures performed with sterile gloves in the context of our study had used nonsterile gloves instead, approximately $660 would have been saved while achieving a similar infection risk outcome. It also must be noted that additional gloves may have been used throughout each procedure for either provider preference or unforeseen glove defects. This would further increase the cost difference between glove options. Similar fiscal analysis has been done for Mohs surgery, which calculated an 850% increase in mean price using sterile gloves over nonsterile gloves, resulting in thousands of dollars of additional cost of annual surgeries.²⁴ Furthermore, research across a dozen environmental impact categories revealed that sterile latex gloves have 11.6 times the burden on climate change as nonsterile gloves, with nonsterile gloves consistently having the lowest impact on the environment across all studied metrics.²⁵ Thus, when examined outside of clinical applications, using nonsterile gloves has proven benefits for both profit and planet.

This study has a few limitations. First, the retrospective nature of the analysis may include selection bias, with medical records potentially yielding incomplete information or unreported infections. Patients lost to follow-up in the postinjection period were also not accounted for to limit attrition bias; however, this may further limit the findings of the study. In addition, there exists some concern for temporal bias given the transition period, although this was mitigated through capturing patients across all seasons both at the time of injection and with subsequent follow-ups. The retrospective approach also limits interpretation of results to being correlated in nature, as causal relationships would require prospective or randomized controlled trials. In addition, the study population came from a single healthcare institution under the practice of a single surgeon, which limits the generalizability of the study to other communities or healthcare settings. Although the studied groups were evaluated to have comparable baseline characteristics, other differences such as history of joint infections were not included in the analysis. The same is true for the immunocompetency, as separate investigation would be required to conclusively extend these results to injections performed on immunocompromised patients. The study's results would be more applicable with a larger study population and a greater or more frequent follow-up occurrence. A larger, multicenter study design would also bolster the robustness of findings. Future studies related to subacromial corticosteroid injections should focus on providing prospective, controlled information, which would further enhance the existing body of literature supporting the comparable postprocedure infection outcomes of nonsterile gloves.

Conclusion

To our knowledge, there is no literature comparing the clinical outcomes and cost between using nonsterile gloves and sterile gloves for corticosteroid shoulder injections. Our results show that despite a substantially higher cost with using sterile gloves compared with nonsterile gloves, there is no difference in infection rate between the two glove groups. This clinical finding combined with the financial and environmental burden of using sterile gloves may push providers into incorporating nonsterile gloves in routine shoulder joint injections.