Joint Efforts: A Review of Large Joint and Bursa Steroid Injections and Aspirations by Primary Care Physicians

Abstract

Large joint and bursa steroid injections and aspirations are useful for many musculoskeletal pathologies commonly seen in primary care. However, there is lack of clarity on whether primary care physicians should perform these procedures themselves or refer to specialists. This study reviews the existing literature to explore the impact of primary care physicians performing these procedures. We performed a PubMed search using two search protocols to identify relevant articles. Two investigators independently screened articles for relevance via specific inclusion and exclusion criteria. Two additional investigators resolved disagreements. We also hand-searched references in all articles. The search methods identified 257 articles, of which 132 met inclusion criteria. Agreement was almost perfect for results from Protocol 1 (kappa = 0.96, 95% CI 0.91–0.99) and substantial for Protocol 2 (kappa = 0.77, 95% CI 0.60–0.94). The results reveal data on primary care physicians performing large joint and bursa steroid injections and aspirations including current trends, management and procedural techniques, outcomes, safety and complication rates, patients’ access to treatments, cost differences between primary care versus subspecialty performed procedures, feasibility concerns, training, and patients’ perspectives. We conclude there is a benefit to primary care physicians performing these procedures with regards to cost, access, and patient satisfaction with no compromise to outcomes, safety, and complication rates.

BACKGROUND

Nearly all musculoskeletal (MSK) pathologies are initially evaluated by primary care physicians (PCPs).¹ MSK pathologies constitute a significant portion of visits and workload in primary care.^2,3 Multiple studies estimate that common pathologies of larger joints such as knees and shoulders often do not resolve even with 6–12 months of conservative management, and 30–40% of cases ultimately undergo steroid injections.^1–3 Large joint and bursa steroid injections and aspirations involving knee, shoulder, or trochanteric bursa—referred hereon as musculoskeletal injection therapy (MIT)—have established clinical utility for multiple pathologies, including osteoarthritis, inflammatory arthritis, bursitis, adhesive capsulitis, and other joint pathologies.^4–9 They serve as safer alternatives to nonsteroidal anti-inflammatory drugs (NSAIDs) and narcotics with fewer complications and contraindications.^1,10–17 Aspirations, as well as response to steroid injections, can aid in the diagnosis of joint pathologies.¹⁸ MITs in the ambulatory setting are most commonly performed on the knee joints, followed by trochanteric bursa and shoulder joints.¹⁹ They remain an important tool among the many therapeutic strategies needed to provide patients a multimodal approach to MSK pain.^12,20

However, PCPs report a lack of clear guidance on when they should perform these procedures themselves or refer to a specialist.²¹ Existing literature does not provide clear guidance. Currently, the majority of these procedures are referred to specialists, mostly Orthopedics and followed by Rheumatology, Sports Medicine, Pain Management, and Physical Medicine and Rehabilitation.^22,23 There are no head-to-head clinical trials comparing the performance and outcomes of such procedures by PCPs versus other specialties.

This review summarizes existing evidence on the impact of PCPs performing MITs. We review the literature on (1) current state of practice, (2) management and techniques of PCPs performing these procedures, (3) outcomes of PCPs performing these procedures, (4) safety and complication rates, (5) patient access to treatments, (6) cost differences between PCPs and subspecialists, (7) feasibility within primary care, (8) training, and (9) patients’ perspectives. These results may help determine if PCPs should perform MITs for their patients.

METHODS

PubMed Search Protocols

To inform this review, in October 2023, we used PubMed to search for relevant articles relating to the performance of MIT in primary care. The search strategy was developed with a research librarian from Harvard Medical School. Two search protocols were used. The first protocol (Protocol 1) included keywords relating to primary care, knee, shoulder, trochanter, and injection (see Appendix). No subspecialties were excluded, and no language or date restrictions were placed.

A second protocol (Protocol 2) was conducted in PubMed to identify articles related to patients’ access to MIT in primary care versus other specialties. A separate protocol was required for this specific topic given healthcare access varies considerably depending on geography and time-period. Unlike Protocol 1 which prioritized a broad search for maximal inclusion of relevant articles, Protocol 2 sought to narrow search results to more current data in the United States (US) and Canada. For Protocol 2, articles were restricted to the last 5 years from 2018 to 2023 and included keywords such as Orthopedics, Sports Medicine, Rheumatology, wait time, and delay of care (see Appendix).

Inclusion and Exclusion Criteria

Articles from Protocol 1 were excluded if they did not discuss the relevant procedures (arthrocentesis or injections of the knee, shoulder, or trochanteric bursa) or if they did not specify the specialty of the proceduralists. Articles from Protocol 2 were excluded if they did not contain relevant information on wait times or delay of care, if they did not discuss the relevant specialties, if they contained only data from outside the US and Canada, or if they were published before 2018. No language restriction was placed, and all study designs were considered.

Two investigators (SM and AM) independently screened titles and abstracts for relevance and reviewed full-length articles when needed. Two additional investigators (BF, IM) resolved disagreements on article inclusion or exclusion. To supplement our search, we conducted a hand search of the references in all articles. Two investigators (SM and AM) manually performed inductive coding to identify recurring themes and topics, which were then reviewed and approved by all coauthors.

RESULTS

Protocol 1 identified 185 articles and Protocol 2 identified 41 articles with no overlap (Fig. 1). Of the 185 articles identified by Protocol 1, 83 met the inclusion criteria. An additional 19 articles found in their references also met the inclusion criteria, totaling 102 studies. Of the 41 identified by Protocol 2, 18 articles met the inclusion criteria. An additional 12 articles in their references also met the inclusion criteria, totaling 30 articles.

Figure 1.

Methodology for selection of articles

Agreement between the two reviewers was almost perfect for results from Protocol 1 (kappa = 0.96, 95% confidence interval 0.91–0.99) and substantial for Protocol 2 (kappa = 0.77, 95% confidence interval 0.60–0.94).

CURRENT STATE OF PRACTICE

MSK presentations accounted for a large portion of primary care appointments.^1–3 Despite this, PCPs widely reported low confidence in MIT skills.²⁴ An American study showed that despite 71% of PCPs routinely suggesting MIT for severe arthritis and bursitis, only 19% would perform it themselves versus 89% who referred patients to specialty clinics.²³ A European study showed a larger portion of PCPs performing MIT (54%), but only a smaller portion (15%) performed more than half of MITs within primary care. Providers who are male and who work in rural localities are more likely to perform MIT.²⁵ The most cited reasons for not performing such procedures are lack of training and comfort in procedural technique and “inability to maintain injection skills.”^23,25 Under-recognition of indications may be another barrier. In one study where PCPs and rheumatologists reviewed the same patient scenarios, rheumatologists recommended joint injections per guidelines more often than PCPs.²⁶ In another study, 20% of patients referred by PCPs to Spine for lower back pain were found instead to have greater trochanteric pain syndrome (GTPS) verified by positive response to a trochanteric bursa steroid injection.²⁷

MANAGEMENT AND TECHNIQUE

Existing literature suggests clinical guidelines and practice patterns regarding MIT do not significantly differ between PCPs and subspecialists. A systematic review of recommendations and guidelines for osteoarthritis found treatment decisions and recommendations for MIT do not differ between primary care and orthopedic guidelines.²⁸ In clinical practice, studies from Canada, the Netherlands, and Ireland show agreement between surgical and nonsurgical providers in their recommendations to patients for MIT versus other treatments.^29–32 Furthermore, an American study surveyed 264 physicians and found no significant differences in the medications and techniques used for steroid injections between PCPs and specialists.³³

OUTCOMES

While there are no head-to-head trials comparing MIT outcomes between PCPs and specialists, many uncontrolled studies have demonstrated positive outcomes of MITs performed by PCPs. Outcomes include short and long-term relief of pain, improved range of motion, reduced disability, better quality of life (QOL), and positive patient satisfaction.^9,34–41 Multiple studies also demonstrate steroid injections have comparable outcomes to physical therapy.^37,42,43 They may even increase patients’ likelihood of engaging in physical therapy.⁴⁴ These procedures are non-inferior to NSAIDs or acupuncture for pain control.^9,45 The outcomes data from all these studies of primary care performed MIT are summarized in Table 1.

Table 1.

Outcomes and Safety of Primary Care Performed Large Joint/Bursa Steroid Injections

Study	Location	Design	Procedure (pathology)	Outcomes	Safety
Bharga et al. 2013	Rochester, MN, USA	Retrospective cohort study of 1188 patients in a primary care clinic	1. Knee steroid injection (N= 598) 2. Shoulder steroid injection (N= 311) 3. Trochanteric bursa steroid injection (N= 279) (for mixed pathologies)	Pain decreased from 3.10 to 2.36 (P= 0.0001) Physical QOL increased from 37.25 to 42.35 (P= 0.0001) Mental QOL increased from 52.81 to 53.54 (P= 0.0001) Patient very or extremely satisfied: 96%
Mantovani Cardoso et al. 2021	Hartford, CT, USA	Retrospective cohort analysis of 132 patients referred to PT, 38 of whom received procedure	Knee steroid injection (for knee osteoarthritis)	Patients receiving procedure were more likely to undergo PT (OR= 1.38, 95% CI 1.14–1.69, rate ratio 2.50, 95% CI 1.82–3.42)
van der Windt et al. 1998	Amsterdam, the Netherlands	RCT of 109 patients receiving procedure (53) versus PT (56)	Shoulder steroid injection (for stiff and painful shoulder)	Improvement of severity: 15 (3 weeks), 26 (7 weeks), and 19 (13 weeks) points better than PT (P= 0.071) Improvement of daytime pain: 12 (3 weeks), 12 (7 weeks), and 9 (13 weeks) points better than PT (P= 0.001) Improvement of nighttime pain: 12 (3 weeks), 14 (7 weeks), and 9 (13 weeks) points better than PT (P= 0.015) Improvement of disability: 13 (3 weeks), 25 (7 weeks), and 10 (13 weeks) points better than PT (P= 0.024) Improvement in external rotation: 9 (3 weeks), 15 (7 weeks), and 9 (26 weeks) points better than PT (P= 0.002) Improvement in abduction: 5 (3 weeks), 5 (7 weeks), and 2 (26 weeks) points better than PT (P= 0.065)	Mild side effects: 47% in procedures versus 44% in PT (most common is pain at injection site, then flushing and menstrual bleeding) Major complications: none
Hay et al. 2003	UK, multiple locations	RCT of 207 patients receiving procedure (104) versus PT (103)	Shoulder steroid injection (for mixed pathologies)	Disability scored improved with procedure by 3.03 at 6 weeks and 4.55 at 6 months and with PT by 2.56 at 6 weeks and 5.97 at 6 months with no significant difference between both groups
Crawshaw et al. 2010	Leeds, England	RCT of 232 patients receiving procedure (115) versus exercise (117)	Shoulder steroid injection (for subacromial impingement syndrome or severe shoulder pain)	Pain reduced by 8.02 (1 week) and 7.73 (6 weeks) more than exercise (P= 0.001), but no significant difference at 24 weeks (P= 0.368) Disability reduced by 5.15 (1 week) and 7.07 (6 weeks) more than exercise (P= 0.001), but no significant difference at 24 weeks (P= 0.380)	Mild side effects: none Major complications: none
Sharma et al. 2016	Bergen, Norway	RCT of 106 patients receiving procedure (36) versus procedure plus saline distension (34) versus usual care (36)	Shoulder steroid injection (for adhesive capsulitis)	Shoulder pain and disability index improved by 20 points more than control (P= 0.001) Numeric pain score reduced by 2 points more than control (P= 0.001) Abduction improved by 8.3 points more than control (P= 0.001) External rotation improved by 10.8 points more than control (P= 0.001) Internal rotation improved by 8.8 points more than control (P= 0.001) Hand behind back improved by 5.0 points more than control (P= 0.001)	Mild side effects: only pain at site and flushing (14%) Major complications: none
Koh et al. 2016	Singapore	Systematic review of 10 RCTs	Shoulder steroid injection (for adhesive capsulitis)	Procedure provides significantly improved pain and functionality versus placebo in all studies, and no statistical difference versus PT or NSAID's at 24 weeks.	Mild side effects: pain after injection (29.8%), facial flushing (12.3%) and menstrual irregularities (10.1%) Major complications: none
Soler-Pérez et al. 2021	Spain, multiple locations	Longitudinal case series of 66 patients receiving procedure for painful shoulder syndrome	Shoulder steroid injection (for osteoarthritis)	Pain (visual analogue scale) reduced from 8.65 to 1.11 (P= 0.001) Pain (constant score) improved from 30.17 to 82.96 (P= 0.001) 22.7% of patients required temporary disability of median 14 days (50% reduction in recovery time)	Mild side effects: 2 patients with flushing (9.4%) Major complications: none
van Doorn et al. 2022	Rotterdam, the Netherlands	Retrospective cohort study of 1000 patients with new shoulder pain and 1000 patients with new shoulder pain receiving procedures	Shoulder steroid injection (for mixed pathologies)	The patient’s age (OR 1.03, 95% CI 1.02–1.04) and a history of shoulder pain (OR 1.52, 95% CI 1.13–2.12) were significantly associated with receiving procedure	Mild side effects: 43 of 1000 patients (4.3%) Most common: local skin reaction, hyperglycemia, and abnormal menstruation Major complications: none
Lievense et al. 2005	Rotterdam, the Netherlands	Retrospective cohort study of 164 patients with GTPS in primary care over 5 years	Trochanteric bursa steroid injection (for GTPS)	Injected patients have 2.7-fold chance of recovery after 5 years as compared with non-injected patients (OR= 0.4, 95% CI 0.1–1.0)
Brinks et al. 2011	Rotterdam, the Netherlands	RCT of 120 patients receiving procedure (60) versus usual care (60)	Trochanteric bursa steroid injection (for GTPS)	Recovery at 3 months seen in 55% of injected patients versus 34% of control (adjusted OR= 2.38, 95% CI 1.14–5.00, NNT= 5) Pain at rest reduced 1.18 points more than control (95% CI 0.31–2.05) Pain with activity reduced 1.30 points more than control (95% CI 0.32–2.29) No significant differences at 12 months	Mild side effects: pain at site < 24 h (38%), no significant difference in hot flushes, menstrual symptoms, or allergic reactions Major complications: none
Smidt et al. 2005	UK and the Netherlands	Heterogeneity analysis of 2 RCTs of 349 patients receiving procedure versus PT versus medications	Elbow steroid injection (for lateral epicondylitis)	Pain during the day and global assessment of change improved in 96% of injected patients versus 59% of non-injected patients (P< 0.05)

SAFETY AND COMPLICATIONS

Studies evaluating MIT performed by PCPs, subspecialists, or comparisons between PCPs and subspecialists all find MIT complications to be rare. The most common complications in decreasing order were pain at the injection site lasting less than 24 h, transient facial flushing, transient irregular menstrual bleeding, rash, transient hyperglycemia, damage to local connective tissue, and needle stick injury to the provider.^5,9,46–48 Serious procedure-related infections ranged between 1 in 2000 and 1 in 50,000, depending on the involved joint, patient co-morbidities, and proceduralists’ experience.⁴⁶

No differences were noted in complication rates of MITs when performed by PCPs versus other specialists. In randomized controlled trials (RCTs) and retrospective cohort studies of PCPs, complication rates and side effects were consistently low and comparable to or even lower than the complication rates reported in the general literature.^{4,5,36–40,47,49} The safety data and complication rates from included studies are summarized in Table 1.

MIT provides PCPs with a safer alternative to oral medications such as NSAIDs or opiates. A systemic review of RCTs on steroid injections for adhesive capsulitis reports that MITs are non-inferior to NSAIDs for pain reduction.⁹ An RCT of PCPs performing steroid injections for shoulder pain showed that patients who received the injection used less oral pain medication at 24 weeks post-procedure.³⁸ A retrospective cohort study of PCP-performed shoulder steroid injections also found patients treated with the procedure used fewer oral medications, and this trend was most notable in elderly patients above the age of 60 years who are at greatest risk for medication-related complications.³¹ Conversely, secondary analysis of three clinical trials for PCPs managing knee osteoarthritis revealed that underutilization of steroid injections correlated with overutilization of oral medications and worse osteoarthritic outcomes such as pain, stiffness, and limited function.⁵⁰

ACCESS TO CARE

A significant shortage of specialists limits access to subspecialty care. US workforce studies reveal a shortage of rheumatologists, reporting insurance and geographic area as major barriers.⁵¹ The American College of Rheumatology membership database reveals predominantly micropolitan areas in the US lacking rheumatologists, with some areas more than 200 miles from the closest rheumatologist.⁵² In an aging population without corresponding increase in rheumatologists, supply models project worsening access issues.⁵³ One study estimates that the number of needed rheumatologists will double that of existing rheumatologists by 2030.⁵¹ Given these shortages, one study found 14% of patients referred to rheumatology and 43% of patients referred to Orthopedics were never actually seen by the specialists.⁵⁴

Such shortages result in long wait times for orthopedic and rheumatology appointments. Wait times ranged from 21 to 199 days for orthopedic appointments and 35 to 78 days for rheumatology appointments, as shown in Table 2.^55–58 In select cases, patients needed to wait more than one year for orthopedic consultation.^32,59,60 Similar patterns were also noted internationally.⁶¹ A lack of standardization or defined criteria for orthopedic triage was cited by one systematic review as a contributor to the health systems’ inability to strategically manage wait times.⁶² From patients’ perspective, wait times for orthopedics and rheumatology consultations were cited as barriers to care.⁶³

Table 2.

Mean and Median Wait Times to Orthopedic and Rheumatology Consultations Following Referral from Primary Care

Study	Data source location	Time of data collection	Specialty	Mean wait time to specialist appointment	Median wait time to specialist appointment
Neimanis et al. 2017	Hamilton, Ontario, Canada	2014	Orthopedics	37.7 days	-
Palani et al. 2023	VHA, USA	2002–2022	Orthopedics	44.0 days	-
Menzies et al. 2011	Arlington, TX, USA	2006–2007	Orthopedics	97 days	-
Thanh et al. 2013	Edmonton, Canada	2009–2011	Orthopedics	132 days	108 days
Penn et al. 2019	VHA, USA	2014–2017	Orthopedics	20.9 days	-
Yee et al. 2022	VHA and Medicare, USA	2014–2017	Orthopedics	24.9 days	-
Shwartz et al. 2023	VHA, USA	2019–2021	Orthopedics	33.4 days	-
Gurewich et al. 2022	VHA, USA	2020–2021	Orthopedics	35.7 days	-
Mohammed et al. 2020	Ontario, Canada	2018–2020	Orthopedics hip	66.3 days	-
Mohammed et al. 2020	Ontario, Canada	2018–2020	Orthopedics knee	74.2 days	-
Vo et al. 2022	Newfoundland and Labrador, Canada	2011–2019	Orthopedics osteoarthritis	84.1 days	73 days
Razmjou et al. 2013	Toronto, Canada	2008–2010	Orthopedics shoulder	198 days	79 days
Menzies et al. 2011	Arlington, TX, USA	2006–2007	Orthopedics spine	199 days	-
Qian et al. 2010	Quebec, Canada	2008	Rheumatology	41.0 days	29 days
Neimanis et al. 2017	Hamilton, Ontario, Canada	2014	Rheumatology	57.9 days	-
Ahluwalia et al. 2020	Ontario, Canada	2015	Rheumatology	-	58 days
Deal et al. 2007	American Medical Association Database, USA	2005–2007	Rheumatology	38 days	-
Thanh et al. 2013	Edmonton, Canada	2009–2011	Rheumatology	77 days	69 days
Katz et al. 2023	Edmonton, Canada	2020–2022	Rheumatology	88.9 days	-
Park et al. 2022	Toronto, Canada	2013–2019	Rheumatology psoriatic arthritis	-	78.5 days

Additionally, long wait times are associated with adverse clinical implications for patients with MSK pain. Patients report decreased QOL during this waiting period,^60,64–66 specifically describing substantial frustration, anxiety, depression, and significant pain.^67,68 Conversely, shorter wait times for MSK interventions resulted in patients’ improved QOL and increased work and employment participation.⁶⁹ Improved pain control was also observed with earlier joint steroid injections.⁷⁰

Significant demographic disparities exist with access to specialty consultation for MSK pain. Patients living in more socially deprived areas reported higher rates of MSK pain,⁷¹ and geographic areas with lower median income correlated with fewer rheumatologists.⁵² Language was also reported as a barrier to rheumatology access.⁵¹ The US Veterans Health Administration (VHA) reports racial disparities with access to orthopedic consultations, with Black and Hispanic veterans experiencing longer wait times than White veterans.⁷² Cross-sectional studies have reported that racial, gender, and age disparities in access to orthopedics and rheumatology consultations have worsened during the post-COVID pandemic era.^73,74

The literature provides examples of primary care–driven MSK management improving specialty access. A national data audit in Ireland revealed only 2% of patients waiting for orthopedic or rheumatology consultations required surgery. An implemented triage program found that 80% of patients initially referred to orthopedics or rheumatology did not require these consultations. By deferring these consultations, wait times for patients who eventually required specialty consultation significantly decreased.⁷⁵ Another single-center intervention in the US found that implementing sports medicine services within primary care resulted in average wait times decreasing from 97 to 19 days for general orthopedics consultation and 199 to 70 days for orthopedic spine consultation.⁷⁶

COST OF CARE

Though many North American health systems do not publicly publish billing charges or reimbursement rates for specific procedures, multiple financial studies suggest significant cost reduction potential with MIT performed within primary care. Firstly, joint steroid injections for arthritis are frequently more cost-efficient than physical therapy, pharmacotherapy, and surgery.^77–82 A meta-analysis of 43 cost-analysis studies found MIT more cost-effective than other pharmacologic treatments for osteoarthritis.⁸³ Secondly, subspecialty referrals require additional office visits, with hospital systems reporting higher billing charges for specialists compared to primary care visits.⁸⁴ The cost of consultation visits can be higher than the cost of large joint steroids injections.⁸⁵

Subspecialty appointments can impose other costs to patients, including delayed treatment, increased reliance on other pharmacotherapies, travel costs for medical appointments, time away from employment due to uncontrolled symptoms, and time away from employment due to additional medical appointments. For example, knee osteoarthritis has been associated with employment reduction.⁸⁶ An Australian study found patients on the waitlist for Orthopedic consultation had a mean economic burden of $40–60 daily from missed or reduced work, 80% of whom ultimately were non-surgical patients.⁸⁷

Performance of MIT by primary care can reduce overall healthcare costs. A multicenter study in the UK found the cost needed to provide PCPs training and medications for shoulder injections was significantly less than the cost reduction associated with a secondary increase in quality-adjusted life years (QALY).⁷⁷ Within the US VHA, providing PCPs training in knee injections yielded an incremental cost-effectiveness ratio of $21,190 per QALY from the VHA perspective and $205 per QALY from the patients’ perspective. Based on Monte Carlo simulations, providing PCPs training in knee injection was cost-effective in 74.4% of simulations for the VHA and 93.6% for patients.⁸⁸ These results support the notion that overall cost for patients and the healthcare system can be reduced when PCPs perform MIT.

FEASIBILITY

Equipment needed for MIT is generally found in most primary care offices, including corticosteroids, local anesthetic agents, needles, syringes, and alcohol pads or other disinfectant solutions. The time required to perform the procedure is performer-dependent but typically short.⁸⁹ While MIT requires sterile technique, they do not require sterile draping or sterile gloves if a no-touch technique is employed. Multiple studies have shown the use of sterile gloves, sterile drapes, masks, and gowns does not provide significant reduction in the already low incidence of procedure-related infections.⁹⁰ As regular gloves cost only $3 per 100 pairs compared to sterile gloves costing $170 per 100 pairs, a no-touch technique may reduce the number of procedural steps, time required for the procedure, and cost burdens to clinics.⁹⁰

With the exception of the hip joint, the benefit of ultrasound guidance for injections of large joints remains uncertain. Current guidelines do not require ultrasound guidance for these procedures.⁹¹ Traditional palpation techniques for MSK injections in larger joints such as the knees, shoulder, and trochanteric bursa are highly effective, with outcomes non-inferior to ultrasound guidance.^92–94

TRAINING

Many PCPs receive training in MIT during their internal or family medicine residency training,^95,96 with 80–90% reporting receiving some education on joint injections and performing at least one procedure during residency.⁹⁷ Simulation-based training is emerging, mostly for the knee and shoulder,^98–106 and can be effective in improving knowledge, skills, and confidence. MSK-focused rotations in community sites with a higher volume of MSK procedures are also available.^107,108 Such clinical experiences have been shown to increase residents’ performance of procedures during residency and post-graduation.¹⁰⁹

However, educational gaps persist. A study from the UK revealed only 22% of joint injections performed by trainees were in the primary care setting versus subspecialty rotations.¹¹⁰ Furthermore, internal medicine residency program directors in the US report the extent of training and mastery of these MSK procedural skills is variable.⁹⁵ Although residents receive didactic and simulation education and observe high volume of MIT in the clinic, few trainees perform more than five procedures on real patients during residency.²³

PATIENT SATISFACTION

MSK symptoms can be debilitating for patients and are associated with significant depression, anxiety, and disability.¹¹¹ A survey study of patients with knee or hip osteoarthritis found 34% of patients endorsed MIT as the most effective and preferred non-surgical treatment option.¹¹² In another study, 57% of those who received knee injections for osteoarthritis found it moderately or extremely helpful.¹¹¹ Among patients receiving knee, shoulder, or trochanteric bursa steroid injections by their PCPs, 76% reported clinical benefit and 96% reported being satisfied or extremely satisfied with the procedural experience.³⁴ Importantly, an inverse relationship between patient satisfaction and wait times was noted.^113–116 The negative impact of long wait times may even be so strong as to negatively impact other unrelated metrics, such as perceptions of information provided by care providers, confidence in the provider, and perceptions of friendliness and concern shown by the provider.¹¹³ Therefore, it may be challenging to assess patient satisfaction without disentangling the impact of wait times.

FUTURE DIRECTIONS

The current literature has yet to answer many questions regarding primary care performed MITs. The impact of multiple clinical factors—different indications for MIT, patients’ medical complexity and comorbidities, and future surgical considerations—on the decision to offer MIT in the primary care setting requires further investigation. Ongoing studies of PCP-performed MITs seek to provide more data on the efficacy, safety, and feasibility of MIT in various clinical scenarios.^117–119 The efficacy and primary care feasibility of newer treatments such as non-steroidal viscosupplementation and platelet-rich plasma injections are also under investigation.¹²⁰ Additionally, optimal implementation of MIT in the primary care setting warrants further exploration. Innovative healthcare delivery models such as dedicated procedure clinics within primary care may offer improved feasibility, provider comfort, and opportunities for education, but the impact on cost and access is unknown.¹²¹ Our literature search did not identify any studies comparing PCPs performing MITs in a dedicated procedure clinic versus their own clinic. Finally, the proper financial structure is needed to support successful implementation of MIT in primary care. Two ongoing RCTs are investigating the cost-effectiveness of joint steroid injections and the most optimal moment for PCPs to refer patients to Orthopedics.^118,119 The equity implications of expanding access to MIT in primary care is yet to be seen but holds incredible promise.

CONCLUSIONS

The evidence summarized in this review highlights the benefits of PCPs performing MIT for patients when clinically indicated. PCPs perform these procedures with good efficacy outcomes and low complication rates that are non-inferior to specialists. When these procedures are performed in the primary care setting, they are more accessible to patients, resulting in lower costs, decreased wait times, and lower burdens of unnecessary referrals. Furthermore, these procedures are safe and feasible in the primary care office, and patients highly appreciate their timely delivery. Despite these benefits, most PCPs have not retained these procedural skills and currently defer these procedures to specialists. Changes in education and attitudes are needed to reverse this trend. In the landscape of increasing MSK pathologies in an aging population, PCPs need to be empowered with a full arsenal of treatment strategies, including large joint and bursa steroid injections and aspirations.

Appendix

The following PubMed search protocols were used to inform this review.

Protocol 1:

((“Access to Primary Care”[Mesh] OR “Primary Health Care”[Mesh] OR “Physicians, Primary Care”[Mesh] OR primary care[tiab] OR primary health care[tiab]) AND (“Shoulder Pain”[Mesh] OR “Shoulder”[Mesh] OR “Shoulder Joint”[Mesh] OR “Knee Joint”[Mesh] OR “Knee”[Mesh] OR trochanter*[tiab] OR knee*[tiab] OR shoulder*[tiab])) AND (“Injections”[Mesh] OR inject*[tiab])

No language or date restrictions were placed.

Protocol 2:

((“Orthopedics”[Mesh] OR “Rheumatology”[Mesh] OR “Sports Medicine”[Mesh] OR “Rheumatologists”[Mesh] OR “Orthopedic Surgeons”[Mesh] OR rheumatolo*[tiab] OR sports medicine[tiab] OR orthopedic*[tiab])) AND (“wait time”[title] OR “delay of care”[title] OR “Care delay”[title] OR “delayed care”[title] OR “waiting lists”[MeSH]) AND (2018:2023[pdat])

No language restrictions were placed, but the date of article publication was restricted to 2018–2023.

Declarations

Conflict of Interest

The authors declare that they do not have a conflict of interest.