Schulthess Clinic Zürich Shoulder Instability Registry: a local registry for multimodal data collection and enhanced care

Abstract

Purpose

The Shoulder Instability Registry (SIR) was established in 2019 to systematically capture and monitor outcomes following surgical treatment of shoulder instability (SI). The aim of this cohort profile is to describe the purpose, design, data structure and baseline characteristics of the SIR, and to outline how the registry supports longitudinal assessment of safety, functional recovery, quality of life and patient-reported outcomes after surgical treatment of SI.

Participants

The registry includes all patients treated surgically for SI. Data collection includes medical history of instability, surgical techniques and intraoperative findings. Clinical assessments include range of motion, instability-specific tests, hyperlaxity signs, Constant Score, subjective shoulder value and SI-specific scores such as the ROWE Score and the Western Ontario Shoulder Instability Index. Radiological evaluations included initial and follow-up imaging via X-rays and CT to assess bony lesions and SI-related arthropathy, as well as MRI for soft tissue injuries. Data are documented preoperatively, at 6 months and at 24 months postoperatively. Although the SIR is an observational cohort rather than a randomised clinical trial, treatment effectiveness is evaluated through longitudinal changes in validated patient-reported outcomes, clinical performance measures and imaging findings.

Findings to date

Between January 2019 and December 2024, 668 patients have been registered (mean age 31 years, 82% men, mean body mass index of 25). According to the American Association of Anesthesiology (ASA) Classification, 66% of patients were classified as ASA I, 33% as ASA II and only 1% as ASA III. 69% of admissions were due to accidents and 31% due to illness. Mean surgery duration was 75 min, and the median hospital stay was 2 days. 38% of patients were insured privately and had general insurance in 62%. 85% of cases were treated arthroscopically, and 15% were treated openly. Baseline clinical scores showed a mean Constant Score of 77 points, mean subjective shoulder value of 49%, mean ROWE Score of 46 points and mean Western Ontario Shoulder Instability Index of 53. Based on Gerber’s classification, 68% of cases were type B2, 29% B3, 2% B5 and fewer than 2% were classified as B4 or B1. 85% of cases suffered from anterior instability, while only 13% experienced posterior instability, the remaining 2% showed multidirectional instability. Among posterior cases, Moroder’s classification identified 58% as type B2, 19% as A2, 7% as A1, 6% as B1, 6% as C1 and 4% as C2. Regarding osteochondral lesions, 20% showed none, 31% showed a glenoid defect, 54% showed a Hill-Sachs lesion and 13% showed a cartilage defect. Scheibel’s classification identified glenoid defects as type 3a in 38% of cases, type 2 in 24%, type 1a in 13% of cases, type 3b in 11%, type 1b in 8% and type 1c in 5% of cases. Positive Gagey and Walch signs were observed in 29% and 27% of cases, respectively. Dislocations presented as primary events in 24% of cases, while 76% were recurrent. Surgical interventions included 459 (70%) Bankart repairs, 6 Bankart plus repairs (<1%) and 108 (16%) Remplissage procedures for soft tissue stabilisation. Bony reconstructions included 52 fragment fixations (8%), 41 coracoid transfers (6%) and 87 iliac crest bone grafts (13%). Additional pathologies were addressed in 533 cases (81%), while 38 cases required revision surgery.

Future plans

We will continue prospectively enrolling and monitoring patients that receive surgical treatment of SI. There are no current plans to halt the data collection in the near future, thereby consistently increasing the number of patients in the registry. A larger availability of data will additionally allow us to apply machine learning modelling and develop risk-prediction tools with the goal of aiding surgical decision making.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• By combining a substantial number of recorded surgical treatments with systematically collected multimodal data, spanning clinical assessments, imaging, intraoperative findings and validated outcome measures—the registry provides a robust and holistic view of recovery patterns and postoperative outcomes.

• Confounders: it is likely that certain confounding variables, which potentially affect the outcome of the treatment, are not collected within the scope of our registry.

• Selection bias: since participants voluntarily contribute to the patient reported-outcome data collection, treatment efficacy might be overestimated.

• External validity: since our clinic is a specialised orthopaedic centre of excellence, the type of treated patients and hence the results of this study might differ in comparison to non-specialist settings.

Introduction

Registries in a clinical setting yield significant potential. They not only serve as data vaults, but also allow for prospective systematic, standardised data collection. By leveraging the large quantities of data collected over time, healthcare providers and scientists are able to validly assess safety and effectiveness of treatments.¹

Shoulder instability occurs when the humeral head moves out of the glenoid socket due to weakened or injured ligaments, tendons or surrounding structures. It is a common condition in active individuals, particularly athletes, and can be classified as anterior, posterior or multidirectional instability (MDI) based on the direction of dislocation. Based on the aetiology of shoulder instability, cases can be classified as either traumatic or atraumatic shoulder instability.²

Traumatic shoulder instability is typically the result of a sudden injury, such as a fall, collision or high-impact activity. This type of instability most commonly occurs in the anterior direction, where the humeral head dislocates forward out of the socket. Traumatic instability often causes damage to soft tissues, such as a Bankart lesion or a Hill-Sachs lesion. It is frequently seen in contact sports or in individuals who have experienced a significant shoulder injury.³

Atraumatic shoulder instability develops gradually and is not associated with a specific injury. It often arises from repetitive stress on the shoulder joint, hyperlaxity (excessive joint looseness) or congenital conditions that predispose individuals to instability. Unlike traumatic instability, atraumatic cases are usually managed conservatively with physical therapy and activity modification.⁴

Posterior shoulder instability is less common but significant, as it involves the humeral head moving backward out of the glenoid. This condition can result from repetitive posterior loading, such as during bench pressing or overhead weightlifting, or from traumatic events like a fall on an outstretched arm.⁵

MDI refers to instability occurring in more than one direction, typically anterior, posterior and inferior. It is often associated with hyperlaxity and is more prevalent in individuals with loose connective tissues or joint conditions such as Ehlers-Danlos syndrome.⁶ In summary, the type of shoulder instability varies depending on the underlying cause and direction of instability.

The treatment of shoulder instability involves both non-surgical and surgical techniques, depending on the severity and type of instability. Non-surgical management is typically the first line of treatment, particularly for atraumatic instability or less severe cases. This approach focuses on physical therapy to strengthen the rotator cuff and scapular stabilisers, improve proprioception and restore joint stability. For cases of recurrent or severe instability, surgical intervention is often necessary. Arthroscopic Bankart repair is commonly performed to reattach the torn labrum and tighten the joint capsule, particularly in anterior instability without significant bone loss.⁷ For patients with substantial glenoid bone loss, either a glenoid augmentation using iliac crest autograft or the Latarjet procedure can be performed, which both provide a bony block to prevent recurrent dislocations. The Latarjet procedure, however, is performed using the short tendon of the biceps and its bony counterpart (coracoid process) to reduce instability.⁸ Remplissage is employed to address engaging Hill-Sachs lesions by filling the humeral head defect with soft tissue to prevent engagement with the glenoid. Other techniques, such as posterior capsular shift, capsular plication or bone grafting, are used for posterior or MDI or when bony defects are present. Emerging approaches, including dynamic stabilisation systems and patient-specific implants, continue to advance treatment options. Both surgical and non-surgical methods aim to restore shoulder function, reduce pain and prevent recurrent instability. The aim of this cohort profile is to describe the purpose, design, data structure and baseline characteristics of the Shoulder Instability Registry (SIR), and to outline how the registry supports longitudinal assessment of safety, functional recovery, quality of life and patient-reported outcomes after surgical treatment of shoulder instability.

Cohort description

The registry is comprised of a retrospective cohort of patients who underwent arthroscopic or open shoulder stabilisation at Schulthess Clinic in Zurich, Switzerland. Data collection for the registry is ongoing, incorporating both retrospective and prospective data. The registry’s completeness is determined by comparing registered cases to all eligible cases in the hospital billing system (congruency rate). The overall congruency rate from January 2019 to December 2024 was 98%.

This consistently high data acquisition can be attributed to three factors: (1) the implementation of a technological solution using FileMaker Pro Advanced (V.21, Claris International, California, USA), which flags each planned shoulder stabilisation procedure for the cohort staff when the operation is scheduled in the electronic clinical record; (2) sufficient funding for study assistance and data management, allowing close monitoring of the operation plan and (3) the high motivation and commitment of the surgeons to quality control and scientific activities. Given the registry’s prospective, continuously enrolling design and the absence of a predetermined stopping point for inclusion, an a priori sample size calculation was deemed not applicable.

Overall, the registry follows a cohort profile structure, capturing why the cohort was established, the populations included, the data elements collected, and how these are followed over time. Specifically, the SIR was created to address the lack of standardised, longitudinal, real-world data on shoulder instability. It systematically collects multimodal information at predefined time points, including patient demographics, clinical examinations, validated Patient Reported Outcome Measures (PROMs), imaging findings, intraoperative data and complications, and monitors patients preoperatively, at 6 months and at 24 months. We report current data completeness and outline the major clinical and research questions the registry is designed to answer, including predictors of recurrent instability, functional recovery after different surgical techniques and the potential for risk-prediction modelling.

In the Swiss insurance system, admissions are classified as either ‘Accident’ or ‘Illness’, where accident insurance covers conditions resulting from a traumatic event and illness insurance covers non-traumatic or degenerative conditions. In this study, this classification was recorded and used as a proxy for distinguishing traumatic from non-traumatic mechanisms of shoulder instability.

Patient and public involvement

Patients were not involved in the design and/or conceptualisation of this study.

Clinical outcomes

Objective shoulder function

At each clinical examination, the following active and passive range of motion parameters are evaluated: flexion, abduction, internal and external rotation at 90° and external rotation at 0°. Range of motion is assessed by the operating surgeon during outpatient visits using a goniometer. Abduction strength of both arms is measured using a spring balance (Pesola, Schindellegi, Switzerland) with the arm in 90° abduction. In addition, several functional clinical tests are performed, such as functional external rotation tests as parts of the Constant-Murley Score (CS), apprehension test, Kim test and Jerk test. Additionally, the Gagey sign, Walch sign and the ROWE Score are assessed.

The ROWE Score

The ROWE Score is an instrument originally intended to measure the success of a Bankart repair after labral tears. It is a 3-item physician-assessed instrument aimed at quantifying shoulder instability, motion and function. The ROWE Score ranges from 0 to 100 points, 100 being the highest attainable value. The ROWE Score has a minimal clinically important difference (MCID) of 9.7 according to the anchor-based method.^{9 10}

Self-report shoulder assessments

Western Ontario Shoulder Instability Index (WOSI)

WOSI¹¹ was developed to reliably and validly measure the disease-specific quality of life in patients suffering from shoulder instability.

The WOSI is a self-reported assessment comprised of 21 items in four subdomains, namely physical symptoms, sports/recreation/work, lifestyle and emotions. The creators of the WOSI paid particular attention to the index having a high responsiveness, which was shown to be greater than that of other commonly used scales such as the disability of the arm, shoulder and hand (DASH), CS or University of California-Los Angeles (UCLA) shoulder rating scale. The minimal important change of the WOSI has been estimated at 14 points.¹²

The subjective shoulder value (SSV)

The SSV is a self-reported measure used to assess a patient’s perceived functional status and pain level of their shoulder. The SSV is often expressed as a percentage of the shoulder’s perceived optimal function. It is an easily administered, reliable, responsive and valid measure of shoulder function in athletes that is highly correlated with other patient-reported outcome measures. The MCID of the SSV is estimated as 13.2.^{13 14}

Quality of life

The EQ-5D is a measure of self-reported health outcomes that is applicable to a wide range of health conditions and treatments. It consists of two parts: a descriptive system (Part I) and a Visual Analogue Scale (VAS) (Part II). Part I of the scale consists of 5 single-item dimensions including: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. Each dimension has a 3-point response scale designed to indicate the level of the problem. Part II uses a vertical graduated VAS to measure health status, ranging from worst imaginable health state to best imaginable health state. The extended version of the EQ-5D includes a valuation task which is used only for valuation studies. There is also an optional set of demographic questions.¹⁵ Prior studies have established an MCID of 0.074 for the EQ-5D instrument.¹⁶

Patient satisfaction

Patients are asked postoperatively to what extent they are satisfied with the outcome of the procedure (0=not satisfied at all; 10=completely satisfied).

Hybrid assessments

CS¹⁷ is one of the most commonly used assessment tools for various shoulder pathologies. Although it is not a gold standard, it has good psychometric properties and has excellent ability to detect clinically significant changes. It assesses four parameters: pain, activities of daily living, strength and mobility. It is considered user-friendly, although the measurement of strength requires some extra effort. A higher score corresponds to better shoulder function. The CS is a relatively unique tool because it combines patient-reported outcomes (pain and activities of daily living: 35 points), performance measurement and clinician-reported outcomes (strength and mobility: 65 points). A distribution-based estimation of the MCID for the CS is 9.4.¹⁸

Adverse event documentation

Adverse events and complications in shoulder instability are evaluated including considerations such as the timing of occurrence (intraoperative or postoperative), the location (local or systemic), whether the event qualifies as a serious adverse event, the affected tissues or physiological systems, potential association with the implant (if any) or with the procedure.

Assessments timeline

Clinical and functional assessments, adverse event assessments and radiological evaluations are conducted pre-operatively and subsequently at 6 months and 2 years after surgery (figure 1). Before each clinical visit, patients are provided with questionnaires assessing patient-reported outcomes. These questionnaires are scheduled and distributed electronically via REDCap. The follow-up rate was 99.3% at baseline, 79.7% at the 6 months and 52.8% at 2 years postoperatively. A decline in follow-up rates over time is expected, especially in younger, more active cohorts, such as this one. Furthermore, since most patients were admitted due to accident, the coverage by the accident insurance often is not given after successful recovery, contributing to the reduced follow-up rates at the 2-year point.

Figure 1. Follow-up timepoints of the Shoulder Instability Registry. AE, adverse event; CS, Constant-Murley Score; FU, follow up; SSV, subjective shoulder value; WOSI, Western Ontario Shoulder Instability Index.

Data integrity and monitoring

All data were captured using REDCap, which is a free, secure, web-based application designed to support data capture for research studies hosted under Schulthess clinic.⁹ Data collection was conducted through an adjunct REDCap to FileMaker Pro method using the SQL Server (see figure 2 for an overview of the registry workflow). FileMaker Pro is a powerful software platform that enables users to create custom applications for managing data. It allows us to design, develop and deploy database solutions tailored to our registry’s specific needs. Participants’ information and study-related data were securely recorded and organised within REDCap’s structured framework. Leveraging SQL Server, REDCap and FileMaker Pro ensures robust data storage, retrieval and management.

Figure 2. Overview of registry workflow. API, Application Programming Interface; CDP, Clinical Data Pool; CRF, Case Report Form; OP-Doku, Surgery Documentation; PROMs, Patient Reported Outcome Measures.

Findings to date

From January 2019 to December 2024, 668 patients have been registered (mean age 31 years, 82% men, mean body mass index of 26). According to the American Association of Anesthesiology (ASA) Classification, 66% of patients were classified as ASA I, 33% as ASA II and only 1% as ASA III. 69% of admissions were due to accidents and 31% due to illness. Mean surgery duration was 75 min, and the median hospital stay was 2 days. 38% of patients were insured privately and had general insurance in 62%. Mean CS, WOSI and SSV at baseline were 77.0 (±13.2), 52.9 (±17.4) and 49.3 (±20.3), respectively. 85% of cases were treated arthroscopically, and 15% were treated openly. Mean preoperative pain was 2.4 (±2.4, scale (0=no pain, 10=maximum pain)), and mean ROWE Score was 45.7 (±22.2). The most common direction of instability was anterior (85%), followed by posterior (13%) and multidirectional (2%); see table 1. The proportion of anterior instability remained quite constant over the years apart from a slight temporary elevation to 93% in 2020.

Table 1. Descriptive characteristics.

Baseline parameter	n	Mean (SD)	Median (range)
Age at surgery	668	30.7 (10.5)	28.8 (14.3–76.4)
Sex, n (%)
Female	119 (18)
Male	549 (82)
Body mass index (BMI)	655	24.6 (3.7)	24.2 (17.1–39.8)
ASA category, n (%)
ASA I: healthy patient	441 (66)
ASA II: mild systemic disease	218 (33)
ASA III: severe systemic disease	8 (1)
Admission type, n (%)
Illness	206 (31)
Accident	462 (69)
Insurance type, n (%)
General	416 (62)
Private	252 (38)
Direction of instability, n (%)
Anterior	534 (85)
Posterior	84 (13)
Multidirectional	13 (2)
Surgery duration (min)	668	74.9 (38.8)	63.0 (17.0–270.0)
Hospital stay (days)	668	2.0 (0.5)	2.0 (0.0–6.0)
Pain NRS	410	2.4 (2.4)	2.0 (0.0–10.0)
CS Constant-Murley score (0=min 100=max)	215	77.0 (13.2)	80.0 (25.1–97.6)
SSV	396	49.3 (20.3)	50.0 (0.0–100.0)
WOSI score	378	52.9 (17.4)	52.7 (7.1–91.0)
ROWE score	369	45.7 (22.2)	45.0 (0.0–100.0)
EQ5D5L utility index DE	396	0.8 (0.2)	0.9 (−0.0 to 1.0)
Abduction	439	146.4 (35.7)	160.0 (10.0–190.0)
Flexion	440	153.8 (30.2)	160.0 (30.0–190.0)
Abduction strength	346	8.2 (3.5)	8.0 (0.0–22.0)

ASA, American Association of Anesthesiology; CS, Constant-Murley Score; EQ5D5L utility index DE, EQ5D5L utility index for German speakers; NRS, Numeric Rating Scale; SSV, subjective shoulder value; WOSI, Western Ontario Shoulder Instability Index.

Strengths

A major strength of the SIR is the systematic, prospective collection of multimodal data in a large cohort of surgically treated shoulder instability patients. The registry integrates demographic characteristics, detailed clinical examinations, validated patient-reported outcome measures, intraoperative findings and standardised imaging classifications at predefined time points. This comprehensive data structure enables a holistic assessment of postoperative recovery, safety and functional outcomes that extends beyond single-score or isolated endpoint analyses.

The registry demonstrates excellent case capture and data completeness, with a congruency rate of 98% when compared with the institutional billing system. This high level of completeness is attributable to the integration of registry workflows into the electronic clinical record, dedicated study personnel and strong engagement of the surgical team. These operational features represent important lessons learnt that may guide the successful implementation of similar registries in other clinical settings.

Another key strength is the longitudinal design, with standardised follow-ups at baseline, 6 months and 24 months, allowing assessment of both short- and mid-term outcomes. The inclusion of instability-specific outcome instruments (eg, WOSI, ROWE Score), in addition to generic measures of function, pain and quality of life, facilitates detailed phenotyping of patient recovery trajectories. As follow-up accrues, the registry is well positioned to support advanced analytical approaches, including risk stratification and machine learning-based modelling.

Finally, the use of standardised and widely accepted classification systems for instability patterns and defects enhances comparability with other registries and published cohorts, supporting external benchmarking and collaborative research.

Limitations

First, the registry is observational and single-centre in nature, which limits causal inference and may affect generalisability. As Schulthess Clinic is a specialised orthopaedic referral centre, the patient population, surgical expertise and treatment strategies may differ from those in non-specialist or lower-volume settings.

Second, despite the high overall data completeness, loss to follow-up increases over time, particularly at the 24 month assessment. This decline is partly attributable to the young, active nature of the cohort and the structure of the Swiss insurance system, in which accident insurance coverage often ends after successful recovery. As a result, long-term outcome analyses may be subject to attrition bias, and treatment efficacy may be overestimated.

Third, while the registry captures a broad range of clinically relevant variables, residual confounding cannot be excluded, as not all potential determinants of outcome (eg, psychosocial factors, adherence to rehabilitation, activity level) are systematically recorded. Additionally, the registry includes only surgically treated patients, limiting comparisons with non-operative management strategies.

Finally, as with any large clinical registry, data accuracy depends on consistent documentation by clinicians and patients. Although robust data monitoring processes are in place, some degree of measurement variability, particularly in clinician-assessed range of motion, cannot be fully eliminated.

Variations in shoulder pathology

85% of cases were treated arthroscopically, and 15% were treated openly. Based on Gerber’s classification, 68% of cases were type B2, 29% B3, 2% B5, and fewer than 2% were classified as B4 or B1. 85% of cases suffered from anterior instability, while only 13% experienced posterior instability, the remaining 2% showed MDI. Among posterior cases, Moroder’s classification¹⁹ identified 58% as type B2, 19% as A2, 7% as A1, 6% as B1, 6% as C1 and 4% as C2. Regarding osteochondral lesions, 20% showed none, 31% showed a glenoid defect, 54% showed a Hill-Sachs lesion and 13% showed a cartilage defect. Scheibel’s classification²⁰ identified glenoid defects as type 3a in 38% of cases, type 2 in 24%, type 1a in 13% of cases, type 3b in 11%, type 1b in 8% and type 1c in 5% of cases. Positive Gagey and Walch signs were observed in 29% and 27% of cases, respectively. Dislocations presented as primary events in 24% of cases, while 76% were recurrent. Surgical interventions included 459 (70%) Bankart repairs, 6 Bankart plus repairs (<1%) and 108 (16%) Remplissage procedures for soft tissue stabilisation. Bony reconstructions included 52 fragment fixations (8%), 41 coracoid transfers (6%) and 87 iliac crest bone grafts (13%). Additional pathologies were addressed in 533 cases (81%), while 38 cases required revision surgery.

Patient-reported outcomes improvement

Recovery, related to both function and pain, as measured by the CS, WOSI, ROWE and SSV demonstrates substantial improvement over time (figures36). Concerning quality of life, improvements are observed over time as well (figure 7). However, the largely high preoperative quality of life, coupled with the ceiling effect, results in a proportionally smaller observed difference between pre- and post-surgical assessments compared with other instruments. When comparing subgroups regarding direction of instability, it seems that postoperative outcomes behave similarly between anterior and posterior dislocations, demonstrating almost identical improvements over time.

Figure 3. Longitudinal results of the constant score. Red line: average constant score across timepoints.

Figure 6. Longitudinal results of the subjective shoulder value. Red line: average subjective shoulder value across timepoints. SSV, subjective shoulder value.

Figure 7. Longitudinal results of the Euroqol Quality of Life Index. Red line: average Euroqol Quality of Life Index across timepoints. QoL, quality of life.

Figure 4. Longitudinal results of the Western Ontario Shoulder Instability Index. Red line: average Western Ontario Shoulder Instability Index (WOSI) across timepoints.

Figure 5. Longitudinal results of the ROWE Score. Red line: average ROWE Score across timepoints.

Adverse events reporting

Adverse events occurred in 35 cases, of which 14 cases experienced redislocations, 6 cases were stiff and 3 cases had persistent pain. Other, more rare adverse events included rotator cuff tears, acromioclavicular joint arthritis, nerve injury, rupture of the biceps, chondrolysis, fracture of the glenoid, hypertrophic keloid scarring, implant migration, labral tears, graft donor site morbidity and thrombophlebitis.

Collaboration

We actively encourage collaboration with researchers in the field of shoulder instability. Data sourced from a multicentric setting allows for a more robust understanding of this heterogenous patient population. The structure of our REDCap database, data collection instruments, the underlying architecture of dataflow, valuable lessons learnt while implementing a new registry in a clinical setting and finally anonymised data can be obtained from the corresponding author on reasonable request. However, a data sharing agreement and ethical approval must be in place prior to such a collaboration.

Epidemiological data across other registries: a comparative overview

The earliest study describing the epidemiology of shoulder instability in a larger population was published in 1984 by Simonet and colleagues. The authors analysed patients who experienced their first traumatic anterior shoulder dislocation and were treated conservatively at the Mayo Clinic, Rochester, Minnesota. They compared patients living in Olmsted County, Minnesota, a quite rural region, to patients referred to the Mayo Clinic from Rochester, a comparatively large urban area. The latter were assumed to be different to the former when viewed from an epidemiological standpoint, since the more severe cases would be more likely to be referred to. They observed that in Olmsted County, two-thirds of patients suffering from traumatic primary anterior shoulder dislocations were men, as opposed to 75.9% in the referral group. Overall, no significant difference could be detected between the groups for age and sex adjusted incidence rates. However, unlike urban-rural differences, sex and age did seem to have a large effect on incidence rates. Specifically, men showed an overall higher incidence rate compared with women, but this didn't hold true for all age categories. In particular, men aged above 60 displayed merely a third of the incidence rate of women; this was irrespective of whether the patients were an Olmsted County or Rochester City resident.²¹

A nationwide epidemiological study was conducted in Italy. Health records were screened for shoulder dislocations requiring hospitalisation over a period of 14 years. The authors collected data from a total of 92 784 hospitalisations due to shoulder dislocation, which corresponds to an incidence rate of 11.2 per 100 000 inhabitants. Results of this study demonstrated a decreasing rate of hospitalisation over the observed time period, exemplifying the continued improvement of outpatient treatment of shoulder dislocations in Italy. The authors reported an overwhelming proportion of male sex in younger age brackets; however, after a transition period between 50 and 60 years of age, female sex takes over the largest proportion of hospitalisations. As a possible reason for this finding, the authors stated the injury mechanism differs greatly between young men and older women, with the former often suffering from athletic injuries, while the latter mostly experience shoulder dislocation from falls.²²

Blomquist and colleagues described the establishment of a Norwegian, multicentre shoulder instability surgery registry and the respective patient outcomes after shoulder stabilisation treatment in the early 2000s. In the first national SIR, the main outcome assessment is the WOSI. However, surgical details are also collected, such as which technique or implant was used or the direction of instability. Depending on the direction of instability, 61%–73% of surgeries were conducted on men. 12 months after surgery, an increase of 16%–24% in the WOSI was observed, based on whether the surgery was primary, and the direction of instability.²³

Another Norwegian study by Liavaag and colleagues analysed medical health records to report the 1-year incidence rate of shoulder dislocations in Oslo, Norway. The authors found that inhabitants of Oslo have a greater incidence rate of shoulder dislocations than other populations previously reported in the literature, which was calculated to be 26.2 per 100 000 inhabitants. Interestingly, the authors also reported a less pronounced increase in incidence rate of shoulder dislocations among older women in comparison to men in the same age bracket.²⁴

A recent study by Smartt and colleagues shines light on the epidemiology of primary anterior instability in older patients. In a cohort of 179 patients aged 50 years or more, who experienced their first anterior shoulder instability between 1994 and 2016, the age and sex adjusted incidence rate was 28.8 per 100 000 person-years.²⁵

The MOON (Multicenter Orthopedic Outcomes Network) shoulder instability study is a prospective multicentre study with 10 participating institutions in the USA. Study assessments are rather extensive, encompassing the American Shoulder and Elbow Surgeons Score, WOSI, the 36-Item Health Survey, the Single Assessment Numeric Evaluation as well as a clinical examination. Men were affected more frequently, with 82% of patients having male sex.²⁶

Owens and colleagues in 2007 described a cohort of students undergoing military training in the US military academy in West Point, New York, USA. During a single academic year, there were 117 traumatic shoulder instability events among 4141 students. 15% of these events were shoulder dislocations, the remaining 85% were classified as subluxations. Men made up 86% of the students that experienced an event and among dislocations, the anterior direction was the most prevalent with 80% of cases. Collectively, this results in an incidence rate of 2.8% (2800 per 100 000 military students). Given the young, male-dominant cohort with a comparatively high physical burden, this rather high incidence rate seems reasonable, even more so when not only considering dislocations but also subluxations.²⁷

Szyluk and colleagues presented the epidemiology of shoulder dislocations in the Polish population by analysing data from electronic health records. They gathered information over a full calendar year for patients diagnosed with post-traumatic shoulder dislocation. The authors stratified the incidence rates of shoulder dislocation by sex, age and place of residence (rural or urban). Overall, there was no significant difference between rural and urban environments, with incidence rates of 25.97 and 26.62 per 100 000 person-years, respectively. The highest incidence rate among subgroups was found in women aged 80+ years, living in urban areas (75.12/100 000 person-years) followed by women aged 70–79 years, living in rural areas (67.40/100 000 person-years). The highest incidence rate among men was found in people living in rural environments aged 60–69 years (62.59/100 000 person-years).²⁸

An investigation into the epidemiology of shoulder dislocations in Aarhus, Denmark, was performed by Kroner and colleagues in 1989. The authors found an incidence rate of 17/100 000 person-years over a 5-year period. Like other authors, they found a bimodal distribution of this injury among age groups. In particular, men aged 21–30 years and women aged 61–80 years experienced the highest incidence rates, which exemplifies the distinct mechanisms of injury with respect to shoulder dislocations among these subcohorts.²⁹ This local registry differs from previously published registries by incorporating standardised imaging classification systems and multimodal assessments, allowing more detailed phenotyping than registries relying primarily on PROMs.

There are numerous epidemiological studies in the realm of shoulder instability. Most of them are based on medical health records, while fewer studies report instability registries that include extensive assessments and include follow-up after treatment (see table 2). While the incidence rate of certain conditions is a very valuable public health metric, it is difficult to draw conclusions when comparing different pathologies, treatments and outcomes. This cohort profile describes the design, methodology and baseline characteristics of the SIR. As follow-up data mature, the registry will enable future analyses on outcomes, risk factors and treatment pathways; however, the present dataset does not permit conclusions about treatment efficacy.

Table 2. Summary of studies describing shoulder instability in a larger cohort.

Author	Location	Year of publication	Data source	Study type	Criterion	Incidence rate
Simonet et al	Minnesota, USA	1984	Medical records	Epidemiological study	Traumatic anterior dislocation	7.5
Kroner et al	Aarhus, Denmark	1989	Medical records	Epidemiological study	Shoulder dislocations	17.0
Owens et al	New York, USA	2007	Prospective data collection/Military academy	Epidemiological study	Traumatic shoulder instability (including subluxation)	2800
Liavaag et al	Oslo, Norway	2011	Medical records	Epidemiological study	Primary shoulder dislocation	26.2
Blomquist et al	Norway	2012	Prospective data collection	Registry	Primary and revision shoulder stabilisation surgery	–
Kraeutler et al	USA	2018	Prospective data collection	Registry	Primary shoulder stabilisation surgery	–
Longo et al	Italy	2021	Medical records	Epidemiological study	Hospitalisation due to shoulder dislocation	11.2
Smartt et al	Minnesota, USA	2022	Medical records	Epidemiological study	Primary anterior dislocation aged 50+	28.2
Szyluk et al	Poland	2022	Medical records	Epidemiological study	Shoulder dislocations	~26.0

The registry includes only surgically treated patients, as consistent long-term follow-up of non-operative cases is not feasible in our setting. This limits comparative analyses between surgical and conservative treatments. With our local SIR, although only providing limited informative value from an epidemiological perspective, modern surgical techniques can be evaluated and prospectively compared in terms of patient outcomes. Additionally, complications and adverse events can be documented and analysed more easily and accurately. The implementation of standardised recording methodologies also facilitates the identification of patterns and risk factors associated with shoulder instability. Although patients were not involved in the initial design of the registry, we plan to incorporate patient input in future updates to enhance its relevance and patient-centred focus.

Data availability statement

Data are available upon reasonable request.