Non-cefazolin antibiotic prophylaxis is associated with increased rates of acute infectious, medical, and surgical complications following primary total shoulder arthroplasty: a database study

Abstract

Hypothesis

This study aimed to evaluate different prophylactic antibiotic regimens and risk for acute postoperative readmissions, infections, and medical complications.

Methods

This retrospective cohort study used the Premier Healthcare Database to identify patients who underwent primary elective shoulder arthroplasty procedures over the age of 18. The database was queried to identify perioperative antibiotic regimens using specific codes for cefazolin, clindamycin, and vancomycin. Patients receiving cefazolin monotherapy for prophylactic antibiotic regimen were used as a control group and multivariate logistic regression was then used to calculate odd ratios for postoperative complications within 90 days of the procedure.

Results

A total of 139,032 patients undergoing primary shoulder arthroplasty between 2016 and 2020 were included. Cefazolin monotherapy was the most prescribed antibiotic regimen (59.3%), followed by vancomycin and cefazolin combination therapy (23.3%), vancomycin monotherapy (8.6%), clindamycin monotherapy (5.6%), and vancomycin and clindamycin combination therapy (3.2%). Monotherapy with clindamycin or vancomycin was associated with increased risk of periprosthetic joint infection (P < .001). Vancomycin use as monotherapy or combination with another antibiotic increased the risk of postoperative acute kidney injury (P = .017-.001).

Conclusion

Compared to vancomycin and clindamycin monotherapy, cefazolin monotherapy was associated with a lower risk of developing acute postoperative periprosthetic joint infection. The results of this study support the use of cefazolin monotherapy as the current standard for patients undergoing primary total shoulder arthroplasty.

Over the past few decades, total shoulder arthroplasty (TSA) has evolved both its techniques and implants to address a growing number of shoulder pathologies including arthritis, avascular necrosis, and rotator cuff injuries.^17,25,26,29 Unsurprisingly, the number of shoulder replacements in the United States has increased drastically from 1995 to 2017.¹⁰ As TSA use continues to rise, orthopedic surgeons must be prepared to handle the ever-increasing need for revision arthroplasty procedures for complications such as component loosening, dislocation, and periprosthetic fracture.¹² Among the most devastating of these complications are periprosthetic joint infections (PJIs), which are estimated to occur in up to 5% of primary cases and remain one of the leading causes for revision arthroplasty.^4,9,27 PJI tends to portend poorer postoperative outcomes, and associated revision procedures can have significant medical, psychological, and financial effects on patients, underscoring the importance of identifying treatments and establishing guidelines to reduce the incidence of PJI following TSA.^2,7

Preventing PJI in TSA remains a difficult task, as the most commonly implicated organism (Cutibacterium acnes) has biological characteristics that make it particularly resistant to many different topical and parenteral treatments.^20,28 As such, research in this area has focused on a variety of approaches, including optimization of modifiable risk factors, preoperative antibiotic prophylaxis, intraoperative surgical techniques, and postoperative wound management.^18,20

Currently there remains no consensus regarding the optimal perioperative antibiotic regimen to prevent infections following shoulder arthroplasty. Although the International Consensus Meeting on Musculoskeletal Infections recommends a weight-based dose of intravenous (IV) cefazolin to patients undergoing shoulder arthroplasty,²⁰ alternative antibiotics such as clindamycin or vancomycin are often used due to penicillin allergy or surgeon preference. Although these alternative antibiotics have gram-positive coverage that should account for C. acnes and common pathogens such as Staphylococcus aureus and Staphylococcus epidermidis, each presents with its own unique pharmacokinetics and pharmacodynamics that may ultimately affect its prophylactic efficacy. Prior studies have compared multiple antibiotic regimens in patients undergoing TSA and demonstrated that both vancomycin and clindamycin prophylaxis may increase rates of PJI and other infectious complications. These studies, however, have either used a single institutional database, compared single antibiotic regimens, or failed to examine how antibiotic choice affects acute readmissions and medical complications.

As such, the purpose of this study was to use a large, multicenter national database to analyze the effectiveness of various antibiotic regimens (including both single and combined regimens) following shoulder arthroplasty. More specifically, we aimed to determine the association between the various antibiotic regimens and acute postoperative PJI, deep and superficial wound infections, wound dehiscence, readmission, and multiple other medical and surgical complications. We hypothesized that treatment with cefazolin alone would result in lower rates of infectious complications and fewer drug-related patient complications as compared to alternative or combined regimens.

Methods

Study population

The study population used for analysis was all adult patients who underwent elective primary TSA surgeries between 2016 and 2020. The study used the Premier Healthcare Database (PHD), which is a nationwide compilation of patient data, procedure/diagnosis codes, and billing records for inpatient medical services. Patient selection was accomplished through the use of International Classification of Disease, 10th Revision (ICD-10) and Current Procedural Terminology codes (eg, 23,472 for anatomic TSA/reverse TSA). Patients below 18 years of age, those who had undergone revision TSA, and those whose primary indication for TSA was proximal humerus fracture, malunion/nonunion after prior fixation of a proximal humerus fracture, or infection were excluded. The study was not subjected to institutional review board evaluation since the PHD does not include confidential patient health information, aligning with the Health Insurance Portability and Accountability Act.

Identification of study cohorts

The database was queried to identify patients undergoing anatomic or reverse TSA. The primary diagnosis associated with surgery was also identified using ICD-10 codes for post-traumatic arthritis, inflammatory arthritis, avascular necrosis, proximal humerus fracture, rotator cuff tear, malunion or nonunion, open reduction and internal fixation complication, and unspecified arthropathy. Once exclusion criteria were applied, perioperative antibiotic regimens were identified using PHD-specific charge codes for cefazolin, clindamycin, and vancomycin; these antibiotics were chosen as they have been shown to be the most common prophylactic agents used for TSA procedures.²⁴ Potential demographic confounders were captured and included age, Charlson Comorbidity Index, length of stay, total cost of care, race, gender, hospital teaching status, hospital size by number of beds, hospital setting, geographic region, and insurance payor type. In addition, common patient comorbid conditions were also recorded.

Analysis

Descriptive statistics were used to present demographic, diagnosis, and comorbidity data. Patients who received on cefazolin as their perioperative antibiotic prophylaxis were used as the control cohort. Patient cohorts identified for comparative analysis were those receiving vancomycin, clindamycin, vancomycin and cefazolin, and vancomycin and clindamycin. Multivariate logistic regression was used to calculate odds ratios (OR) for the outcomes of periprosthetic joint infection (PJI), wound dehiscence, deep wound infection, superficial infection, periprosthetic fracture, dislocation, and prosthetic joint loosening while controlling for confounding demographic and comorbidity variables. ORs for 90 day hospital readmission and multiple medical complications, including acute kidney injury (AKI), acute respiratory failure, deep venous thrombosis (DVT), myocardial infarction, pneumonia, sepsis, and stroke, were also calculated in an identical fashion. Statistical analyses were performed in STATA (version 17.0; StataCorp, College Station, TX, USA). Descriptive characteristics were analyzed for mean and standard deviation. Missing entries were substantially less than 2.5% for each variable. For all analyses, statistical significance was defined as an alpha of 0.05.

Results

Our analysis consisted of a total of 139,032 patients undergoing primary shoulder arthroplasty over the course of 5 years (2016-2020); patient demographics and comorbidities are listed in Table I and Table II, respectively. Cefazolin monotherapy was the most commonly prescribed perioperative antibiotic, used in 77,109 (59.3%) patients followed by a dual therapy of vancomycin and cefazolin (33,274; 23.3%), vancomycin monotherapy (11,189; 8.6%), clindamycin monotherapy (7284; 5.6%), and vancomycin and clindamycin dual therapy (4112; 3.2%). The use of vancomycin monotherapy has decreased over time from 9.43% of all antibiotics administered in 2016 to 6.62% in 2020. A similar trend was observed for clindamycin with its utilization decreased from 6.85% in 2016 to 3.72% in 2020. In comparison, dual therapy consisting of vancomycin and cefazolin increased in use rates from 15.43% in 2016 to 28.72% in 2020 (Fig. 1).

Table I.

Baseline patient demographics.

Variable	Cefazolin (N = 77,132)	%	Vancomycin (N = 11,191)	%	Clindamycin (N = 7287)	%	Cefazolin + vancomycin (N = 30,277)	%	Vancomycin + clindamycin (N = 4113)	%	P value
Age (yr)	69.5		69.5		69.8		69.3		69.0		<.001
CCI (average)	3.4		3.5		3.5		3.4		3.5		<.001
LOS (d)	1.5		1.7		1.7		1.5		1.7		<.001
Total cost (USD)	17,698		18,849		18,110		18,479		18,698		<.001
Male	35,224	45.7	4002	35.8	2409	33.2	13,672	45.2	1462	35.6	<.001
Race											<.001
Asian	3468	4.5	445	4.0	292	4.0	1370	4.5	334	8.1
Black	4035	5.2	503	4.5	327	4.5	1543	5.1	209	5.1
Caucasian	68,232	89.0	10,061	90.3	6554	90.4	26,939	89.3	3537	86.2
Other	958	1.3	134	1.2	80	1.1	316	1.1	25	0.6
Hispanic (%)	3.4		2.5		3.2		2.6		3.1		<.001
Payer category											<.001
Private	16,468	27.0	2223	25.1	1458	24.9	6667	27.6	919	28.1
Medicare	40,044	65.6	5946	67.1	3995	68.1	15,666	64.9	2145	65.5
Medicaid	2555	4.2	439	5.0	236	4.0	1026	4.3	122	3.7
Workers' comp	1760	2.9	215	2.4	146	2.5	671	2.8	78	2.4
Other	222	0.3	37	0.4	31	0.5	94	0.4	11	0.3
Bed size											<.001
<99	7359	9.6	820	7.3	633	8.7	3190	10.5	386	9.4
100-199	13,832	17.9	1997	17.9	1434	19.7	5250	17.3	1294	31.5
200-299	12,550	16.3	2056	18.4	1196	16.4	6132	20.3	807	19.6
300-399	12,894	16.7	1523	13.6	923	12.7	4165	13.8	519	12.6
400-499	9367	12.1	1012	9.1	1180	16.2	4405	14.6	460	11.2
>500	21,094	27.4	3772	33.7	1919	26.3	7117	23.5	645	15.7
Setting											<.001
Urban	67,421	87.5	9894	88.5	6154	84.5	27,340	90.4	3438	83.6
Rural	9675	12.5	1286	11.5	1131	15.5	2919	9.6	673	16.4
Teaching											<.001
Yes	42,788	55.5	6677	59.7	4427	60.8	17,925	59.2	2800	68.1
No	34,344	44.5	4514	40.3	2860	39.2	12,352	40.8	1313	31.9
Region											<.001
Midwest	22,068	28.6	2337	20.9	2116	29.1	7992	26.4	1133	27.5
Northeast	9434	12.2	1973	17.6	788	10.8	3670	12.1	284	6.9
South	34,282	44.5	5106	45.7	3744	51.4	14,662	48.5	2383	58.0
West	11,312	14.7	1764	15.8	637	8.7	3935	13.0	311	7.6

CCI, Charlson Comorbidity Index; LOS, length of stay; USD, United States dollar.

Table II.

Patient comorbidities.

Variable	Cefazolin (N = 77,132)	%	Vancomycin (N = 11,191)	%	Clindamycin (N = 7287)	%	Cefazolin + vancomycin (N = 30,277)	%	Vancomycin + clindamycin (N = 4113)	%	P value
CHF	4011	5.2	672	6.0	428	5.9	1524	5.0	226	5.5	<.001
CKD	6363	8.3	1054	9.4	627	8.6	2476	8.2	374	9.1	<.001
Coagulopathy	1152	1.5	202	1.8	130	1.8	487	1.6	61	1.5	.05
Collagen disorder	4273	5.5	795	7.1	495	6.8	1879	6.2	295	7.2	<.001
COPD	14,502	18.8	2722	24.3	1772	24.3	5887	19.4	968	23.5	<.001
Iron deficiency anemia	1037	1.3	181	1.6	115	1.6	366	1.2	64	1.6	.006
DM (uncomplicated)	11,867	15.4	1773	15.8	1253	17.2	4639	15.3	647	15.7	.001
DM (complicated)	5330	6.9	903	8.1	573	7.9	2253	7.4	325	7.9	<.001
HIV/AIDS	43	0.1	6	0.1	4	0.1	31	0.1	1	0.02	.07
HTN (uncomplicated)	45,397	58.9	6641	59.3	4344	59.6	17,944	59.3	2469	60.0	.32
HTN (complicated)	8091	10.5	1291	11.5	832	11.4	3156	10.4	457	11.1	.001
Hypothyroidism	13,129	17.0	2316	20.1	1544	21.2	5392	17.8	829	20.2	<.001
Liver disease	1188	1.5	222	2.0	108	1.5	528	1.7	62	1.5	.002
Obesity	16,361	21.2	2535	22.7	1523	20.9	6750	22.3	979	23.8	<.001
PVD	2528	3.3	383	3.4	231	3.2	979	3.2	136	3.3	.88
Pulmonary disorder	713	0.9	123	1.1	74	1.0	293	1.0	47	1.1	.29
Valve disease	670	0.9	105	0.9	56	0.8	266	0.9	39	1.0	.78

CHF, congestive heart failure; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; HIV, human immunodeficiency virus; AIDS, acquired immunodeficiency syndrome; HTN, hypertension; PVD, peripheral vascular disease.

Figure 1.

Yearly Trends of the most common perioperative antibiotic regiment for TSA. TSA, total shoulder arthroplasty.

Vancomycin monotherapy and clindamycin monotherapy were associated with an increased risk of acute PJI within 90 days of operation (OR = 1.90; 95% confidence interval [CI] 1.37-2.64; P < .001 and OR = 2.2; 95% CI 1.51-3.19; P < .001, respectively) compared with cefazolin monotherapy (Table III). Vancomycin monotherapy had a statistically significant association with increased risk for several complications within 90 days of operation including readmission (OR = 1.15; 95% CI 1.03-1.32; P = .016), deep vein thrombosis (DVT; OR = 1.67; 95% CI 1.13-2.49; P = .011), pneumonia (1.50; 95% CI 1.20-1.87; P < .001), acute respiratory failure (OR = 1.22; 95% CI 1.04-1.44; P = .018), periprosthetic fracture (OR = 1.46; 95% CI 1.10-1.94; P = .009), and dislocation (OR = 1.36; 95% CI 1.09-1.77; P = .021).

Table III.

Multivariate model for acute postoperative complications.

	Reference group: cefazolin alone
	Vancomycin alone	Clindamycin alone	Vancomycin + cefazolin	Vancomycin + clindamycin
Hospital readmission	OR 1.16 95% CI 1.02-1.32 P = .02	OR 1.15 95% CI 0.99-1.33 P = .07	OR 0.95 95% CI 0.87-1.04 P = .31	OR 1.03 95% CI 0.84-1.26 P = .78
PJI	OR 1.90 95% CI 1.37-2.65 P < .001	OR 2.19 95% CI 1.51-3.19 P < .001	OR 1.01 95% CI 0.77-1.33 P = .94	OR 1.55 95% CI 0.88-2.72 P = .13
Sepsis	OR 1.63 95% CI 1.28-2.10 P < .001	OR 1.08 95% CI 0.76-1.54 P = .66	OR 0.94 95% CI 0.76-1.16 P = .59	OR 1.49 95% CI 1.00-2.21 P = .05
Superficial wound infection	OR 0.85 95% CI 0.30-2.40 P = .75	OR 1.29 95% CI 0.45-3.67 P = .64	OR 0.88 95% CI 0.44-1.75 P = .71	OR 0.59 95% CI 0.08-4.36 P = .61
Deep wound infection	OR 1.28 95% CI 0.72-2.28 P = .40	OR 1.03 95% CI 0.47-2.24 P = .94	OR 1.06 95% CI 0.70-1.62 P = .77	OR 0.74 95% CI 0.23-2.34 P = .60
AKI	OR 1.27 95% CI 1.10-1.47 P = .001	OR 1.05 95% CI 0.87-1.28 P = .61	OR 1.14 95% CI 1.02-1.27 P = .02	OR 1.17 95% CI 0.93-1.49 P = .18
DVT	OR 1.65 95% CI 1.11-2.45 P = .01	OR 1.30 95% CI 0.77-2.19 P = .33	OR 0.73 95% CI 0.51-1.06 P = .10	OR 0.43 95% CI 0.14-1.36 P = .15
PE	OR 1.28 95% CI 0.87-1.90 P = .21	OR 1.04 95% CI 0.62-1.74 P = .88	OR 0.80 95% CI 0.58-1.10 P = .17	OR 0.82 95% CI 0.38-1.74 P = .60
Pneumonia	OR 1.48 95% CI 1.19-1.84 P < .001	OR 0.99 95% CI 0.72-1.36 P = .95	OR 1.10 95% CI 0.93-1.31 P = .27	OR 1.07 95% CI 0.72-1.60 P = .73
Acute respiratory failure	OR 1.20 95% CI 1.02-1.41 P < .001	OR 1.16 95% CI 0.95-1.42 P = .15	OR 0.93 95% CI 0.81-1.05 P = .25	OR 0.95 95% CI 0.72-1.27 P = .75
MI	OR 0.84 95% CI 0.53-1.33 P = .47	OR 0.89 95% CI 0.52-1.55 P = .69	OR 0.72 95% CI 0.52-1.00 P = .05	OR 1.04 95% CI 0.53-2.03 P = .92
Stroke	OR 0.73 95% CI 0.42-1.28 P = .27	OR 1.43 95% CI 0.87-2.34 P = .16	OR 0.84 95% CI 0.59-1.18 P = .31	OR 0.86 95% CI 0.38-1.96 P = .73
Hematoma	OR 1.23 95% CI 0.83-1.81 P = .31	OR 0.91 95% CI 0.53-1.57 P = .73	OR 0.87 95% CI 0.65-1.17 P = .35	OR 1.62 95% CI 0.94-2.80 P = .09
Seroma	OR 1.58 95% CI 0.70-3.59 P = .27	OR 3.28 95% CI 1.56-6.90 P = .002	OR 0.85 95% CI 0.43-1.67 P = .63	OR 2.46 95% CI 0.86-7.00 P = .09
Periprosthetic fracture	OR 1.46 95% CI 1.10-1.94 P = .008	OR 0.97 95% CI 0.65-1.44 P = .89	OR 1.15 95% CI 0.92-1.43 P = .22	OR 1.30 95% CI 0.82-2.05 P = .27
Dislocation	OR 1.36 95% CI 1.04-1.76 P = .023	OR 1.16 95% CI 0.82-1.63 P = .41	OR 0.96 95% CI 0.79-1.17 P = .67	OR 1.14 95% CI 0.73-1.78 P = .55

PJI, periprosthetic joint infection; AKI, acute kidney injury; DVT, deep venous thrombosis; PE, pulmonary embolism; MI, myocardial infarction.

Bold indicates statistical significance.

Vancomycin, either in monotherapy or dual therapy with cefazolin, was also associated with higher rates of AKI within 90 days of operation (OR = 1.28; 95% CI 1.10-1.48; P = .001 and 1.14; 95% CI 1.02-1.26; P = .017, respectively). There was a similar trend with monotherapy vancomycin use or in combination with clindamycin and an increased risk for postoperative sepsis (OR = 1.64; 95% CI 1.28-2.09; P < .001 and OR = 1.49; 95% CI 1.00-2.21; P = .05, respectively).

There was no statistically significant association between the use of vancomycin monotherapy or in combination with clindamycin or cefazolin and risk for postoperative wound dehiscence, deep wound infection, superficial infection, pulmonary embolism, stroke, hematoma, or prosthetic joint loosening (P = .05-.94).

Discussion

The present study investigated the association between preoperative prophylactic antibiotic regimens in primary TSA and postoperative adverse outcomes. Notably, vancomycin monotherapy and clindamycin monotherapy were associated with increased risk of acute PJI as compared to the control group of cefazolin monotherapy. Furthermore, vancomycin monotherapy was associated with an increased risk of readmission, DVT, pneumonia, and acute respiratory failure. Vancomycin use, either monotherapy or in combination with cefazolin, was associated with higher rates of AKI. Sepsis rates were also higher among patients receiving either vancomycin monotherapy or vancomycin and clindamycin dual therapy. In addition, although the majority of patients in our cohort received IV cefazolin monotherapy as antibiotic prophylaxis in 2020, rates of cefazolin monotherapy are decreasing. Conversely, rates of cefazolin/vancomycin dual therapy rates are increasing.

PJI is one of the most feared and common complications of total joint replacement operations.⁹ As such, strategies to mitigate the risk of postoperative infection in the arthroplasty patient are of significant interest. There is a substantial base of literature regarding choice of antibiotic prophylaxis in the total hip arthroplasty and total knee arthroplasty literature; however, data on this subject among TSA patients are more limited.^1,6,15,23 One study used the Swedish Shoulder Arthroplasty Register to analyze differences in reoperation rates for infection following primary shoulder arthroplasty.⁸ The authors found that cloxacillin monotherapy was associated with a higher risk of reoperation for infection than cloxacillin/benzylpenicillin dual therapy or clindamycin monotherapy.⁸ Notably, this study did not include cefazolin, which may limit its generalizability and applicability to an American cohort of patients in which cefazolin monotherapy is the current standard of care. Marigi et al. found that PJI-free survivorship at 1 month, 1 year, and 2 years was greater in shoulder replacement operations in which cefazolin monotherapy was administered.¹⁸ As in the present study, vancomycin and clindamycin monotherapy were associated with a higher risk for PJI.¹⁸ Thus, our study adds to the existing but limited literature linking alternative therapies with increased PJI risk.

This increased risk with vancomycin monotherapy may be attributable to vancomycin not having gram-negative coverage.^13,15 In addition, vancomycin may be underdosed in certain patients, leading to inadequate antibiotic prophylaxis.^15,19 For example, one study found that incomplete preoperative dosing of vancomycin prophylaxis for primary shoulder arthroplasty was associated with higher rates of infectious complications and reoperation compared to patients whose antibiotic infusions were completed at least 30 minutes before incision.¹⁹ Similarly, it is possible that the method of delivery of vancomycin may be related to its efficacy in preventing infectious complications in TSA. Garofalo et al. studied 827 patients undergoing TSA. Of 422 patients who received intrawound vancomycin powder, zero infections were observed.¹¹ Conversely, of the 405 patients analyzed in their cohort who did not receive intrawound vancomycin, 13 (3.2%) infectious complications occurred.¹¹ Recent basic science studies have further demonstrated the promise of intrawound vancomycin for prevention of C. acnes infection.^14,21 As such, it is possible that intrawound vancomycin powder may be a beneficial adjunct to IV cefazolin, although further investigation is needed.

Although the present study demonstrated increased risk of PJI associated with vancomycin and clindamycin monotherapy prophylaxis and increased rates of adverse medical outcomes associated with vancomycin use, surgeons may still be tempted to avoid cefazolin administration in patients with listed penicillin allergies. Multiple recent studies, however, have suggested overall benefit of cefazolin use in this setting. Kurcz et al. found no statistically significant difference in the rate of allergic reactions to prophylactic antibiotics among those with and without listed penicillin allergies in their cohort of 2451 total joint arthroplasty patients.¹⁶ In addition, patients receiving cefazolin had a lower rate of PJI.¹⁶ Bahoravitch et al. found that patients with reported penicillin allergies had higher risk of revision due to PJI at 30 days, 90 days, 1 year, and 2 years among a large cohort of TSA patients, perhaps owing to alternative antibiotic use.³ Furthermore, Norvell performed a chart review of patients with a labeled penicillin or cephalosporin allergy undergoing total knee arthroplasty or total hip arthroplasty.²² Patients were given either cefazolin or vancomycin and/or clindamycin. The authors found no difference in intraoperative hypersensitivity reactions between the 2 groups and found that patients in the cefazolin group had fewer PJIs and skin and soft tissues infections.²² As such, there is growing literature to support the use of cefazolin as antibiotic prophylaxis, even in those with labeled penicillin or cephalosporin allergies.

Notably, one recent study conducted by Bragg et al. analyzed the economic ramifications of preoperative allergy testing for patients with reported penicillin or cephalosporin allergies.⁵ The authors found that preoperative allergy testing would be economically justified in TSA patients if it prevents one infection out of 223 operations.⁵ These data speak to the significant economic and medical impact of infectious complications following TSA.

In addition to the increased risk of PJI seen with vancomycin and clindamycin monotherapy, higher rates of hospital readmission, AKI, DVT, pneumonia, sepsis, and acute respiratory failure were observed with vancomycin monotherapy. Although our multivariate analysis demonstrated that vancomyin was independently associated with these complications, it is possible that these increased rates may be related to surgeons choosing vancomycin for patients who are more medically complex and thus at greater risk for postoperative complications. Further investigation is warranted to better understand the criteria on which surgeons base decisions about antibiotic prophylaxis, and how alternative antibiotic regimens may affect long-term outcomes beyond 90 days after surgery.

The present study has inherent strengths, including a large patient cohort with data from multiple medical centers, analysis of multiple combinations of antibiotic prophylaxis, and a thorough multivariate analysis that accounted for multiple patient demographic factors and comorbidities. However, there were also multiple limitations present. The results of this study were dependent on proper ICD-10 and Current Procedural Terminology coding, which leaves room for inaccurate coding or misclassification. Similarly, given the retrospective nature of the study, we were unable to infer causative relationships between antibiotic prophylaxis and the outcomes measured. The database used in this study did not have granular data regarding why certain patients received alternative antibiotic regimens, which would provide valuable information regarding the decision-making process of the surgical team. Finally, the present study only evaluated short-term outcomes. Future studies may seek to better understand when and why surgeons may opt for alternative antibiotic regimens in TSA.

Conclusion

Cefazolin monotherapy was associated with a lower risk of acute PJI in primary TSA compared with clindamycin and vancomycin monotherapies. Furthermore, vancomycin monotherapy was associated with an increased risk of readmission, DVT, pneumonia, AKI, and acute respiratory failure. Although cefazolin monotherapy remains the most prevalent choice for antibiotic prophylaxis in shoulder replacement surgery, rates of this antibiotic regimen are decreasing. In order to minimize adverse patient outcomes, cefazolin monotherapy should remain the gold standard for preoperative antibiotic prophylaxis before TSA.

Disclaimers

Funding: No funding was disclosed by the authors.

Conflicts of interests: Frank A Petrigliano reports a relationship with Exactech Inc. that includes consulting or advisory and a relationship with Stryker Orthopaedics that includes consulting or advisory. Joseph N Liu reports a relationship with Stryker Orthopaedics that includes speaking and lecture fees, a relationship with Innocoll Biotherapeutics NA Inc. that includes travel reimbursement, a relationship with Arthroscopy Association of North America as a board or committee member, and a relationship with American Shoulder and Elbow Surgeons as a board or committee member. The other authors, their immediate families, and any research foundation with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.