Missed postoperative orthopaedic outpatient follow-up has been associated not only with patient dissatisfaction and difficulty in timely identification of potential complications but also with provider dissatisfaction, wasted healthcare resources, and disruption of clinical research efforts.1,2 In addition to its effect on assessing clinical outcomes, being lost to follow-up (LTFU) can negatively affect statistical power and represents a possible source of bias in research.3 In a systematic review by Somerson et al,4 the authors evaluated 448 clinical orthopaedic studies that reported on LTFU. They reported a mean LTFU rate in orthopaedic randomized controlled trials of 10.4%, with studies having longer follow-up periods and those performed in the United States at higher risk. In a multicenter study evaluating LTFU after anterior cruciate ligament reconstruction, with LTFU defined as failure to complete 2-year patient reported outcomes, the authors determined an 88% follow-up rate. Independent risk factors of increased LTFU included male sex and non-White race. Another randomized controlled trial of anterior cruciate ligament reconstruction with or without lateral extra-articular tenodesis reported a similar LTFU rate, although the independent risk factors included current or previous smokers, those employed part-time, and those with body mass index ≥25 kg/m2. Most of the existing orthopaedic literature regarding LTFU has focused on total hip and knee arthroplasty,5–7 trauma,3,8,9 hand fractures,10 and hand/upper extremity surgery11 with fewer studies evaluating the prevalence and risk factors of LTFU after shoulder arthroplasty12 and arthroscopic rotator cuff repair.13

The effect of LTFU on the clinical outcomes remains difficult to generalize. Within the arthroplasty literature, LTFU has been associated with poor surgical outcomes, missed complications, and increased need for revision surgeries.14,15 Murray et al14 reported a 20% LTFU rate at 15 years after total hip arthroplasty and concluded that patients who missed follow-up had markedly worse hip pain, motion, and assessment of their surgical outcomes. By contrast, Samade et al11 reported that although 26% of their cohort of upper extremity postoperative patients was LTFU, the median satisfaction score was 10/10 when contacted by telephone. Clohisey et al5 reported that one-time verbal instructions to return for follow-up resulted in poor follow-up attendance, whereas others have postulated as to the benefits of automated phone call reminders before the scheduled visit.11 Additional strategies have been suggested to minimize the LTFU rate, including the implementation of telemedicine to improve patient compliance. In a pilot study by Fabrés Martín et al,16 virtual follow-up visits were associated with reduced visit times, reliably assessed range of motion, and maintained satisfaction perception for patients after total knee and reverse shoulder arthroplasty. O'Donnell et al17 evaluated the effectiveness and patient satisfaction of virtual visits after shoulder surgery, including rotator cuff repair and both anatomic total shoulder arthroplasty and reverse total shoulder arthroplasty. When compared with in-person visits, the virtual visits were associated with decreased cost, shorter visit time, and high patient satisfaction. Understanding which patients are at highest risk of noncompliance with scheduled postoperative visits would be useful to focus additional resources on maximizing their clinical outcome and minimizing the risk of missed adverse events.

The purpose of our study, therefore, was to describe the prevalence and risk factors of missed short-term follow-up (STFU) after elective arthroscopic rotator cuff–related surgery and to identify any patient-specific or surgical-related parameters that were predictive of missed follow-up at the 6 and 12-month postoperative visits. Furthermore, we sought to determine whether the implementation of a standardized electronic patient contact system would decrease the risk of missed follow-up. We hypothesized that patients with more deprived socioeconomic status, notable medical comorbidities, greater distance from the orthopaedic clinic, and a pattern of previously missed outpatient visits would be more likely to miss the 6 and 12-month follow-ups, whereas those who presented with reminders of their upcoming visit would be more consistent with their scheduled follow-ups.

We conducted a retrospective review of all the elective arthroscopic shoulder procedures performed at our tertiary referral center between February 2016 and January 2022 by a single fellowship-trained shoulder surgeon. Inclusion criteria consisted of all consecutive patients who underwent primary or revision arthroscopic rotator cuff surgery with a minimum of 12-month follow-up eligibility. We excluded any patient who underwent shoulder arthroscopy exclusively for non–rotator cuff–related pathology (ie, isolated distal clavicle resection, biceps tenodesis, instability repair), any patient who died within the 12-month follow-up period, or any patient who explicitly stated that they could no longer return for follow-up because they were permanently leaving the area. Furthermore, any patient with a scheduled postoperative visit that fell between March 2020 and July 2020 was excluded in light of the COVID pandemic. No patients were contacted specifically for follow-up after initiation of the study apart from the standardized electronic patient appointment reminders. The study was approved by our institutional review board (2022-14454).

We reviewed the electronic medical records to collect demographic data including patient age, sex, body mass index, marital status, preferred language (Spanish versus English versus other), insurance type (commercial insurance versus Medicare versus Medicaid versus workers' compensation), self-identified race, smoking status (current versus former versus never smoked), documented history of depression and/or anxiety, employment status, straight-line distance from home to clinic, and age-adjusted Charlson Comorbidity Index. The age-adjusted Charlson Comorbidity Index is a validated instrument for predicting mortality based on comorbid disease and was a modification of the original Charlson Comorbidity Index by adjusting risk by age.18 A patient's socioeconomic status was determined by use of the Area Deprivation Index (ADI), an instrument that allows for ranking of neighborhoods by socioeconomic disadvantage at both the state and national levels. The ADI is a validated, multidimensional tool based on multiple factors including income, education, and employment.19,20 At the state level, the data are categorized by deciles (1-10), with a score of 1 indicating the least disadvantaged and 10 being the most disadvantaged. At the national level, the data are given as percentiles, with lower values representing the least disadvantaged. We analyzed ADI data both continuously and in quartiles to compare the most disadvantaged quartile with the least disadvantaged quartile.

Surgical data included nature of concomitant procedures (ie, use of allograft, biceps tenodesis, capsular release), American Society of Anesthesiologists score (1 or 2 versus ≥3), number of anchors used, history of prior ipsilateral or contralateral nonarthroscopic shoulder surgery, prior ipsilateral or contralateral arthroscopic shoulder surgery, subsequent upper or lower extremity arthroscopy, and any ipsilateral revision surgery. The number of anchors placed was used as a surrogate for rotator cuff tear size based on the findings of Shah et al.21 Procedures were categorized as primary repair of the rotator cuff with or without decompression/débridement (type 1) versus those with concomitant use of allograft (superior capsular reconstruction, lower trapezial tendon transfer, collagen bioinductive allograft augmentation), biceps tenodesis, or distal clavicle excision (type 2). Regarding postoperative follow-up visits, patients were routinely scheduled at 1, 3, 6 weeks, 3, 6, and 12 months. For the purpose of this study, we used the following range definitions for each visit point: 3 ± 1 week, 6 ± 1 week, 3 ± 1 month, 6 ± 1 month, and 1 year ± 6 weeks. The outcome measure was appearance at the 6 and 12-month follow-up visits, and nonattendance at those visits was denoted as missed STFU. A patient was deemed eligible for the 12-month follow-up if the time between their surgical date and the date of our final database access surpassed 10.5 months, the lower end of our definition for 12-month follow-up in this study. Any patient who was eligible for and missed the 12-month follow-up, and who never returned for another outpatient evaluation with the treating surgeon, was deemed LTFU to the treating surgeon. Finally, we ascertained whether a patient was treated by another provider within our group for an unrelated musculoskeletal report after being considered LTFU to the treating shoulder surgeon. A patient fulfilling this criterion was denoted as LTFU to the practice because these patients could potentially have been identified as LTFU from the treating shoulder surgeon and redirected for follow-up. The period of follow-up was determined from the date of surgery to the date that our institutional database was last accessed. In January 2018, several electronic patient contact tools were implemented by our orthopaedic department which included formal electronic patient messaging (ie, automated phone calls, alerts through patient electronic medical records) to remind patients of upcoming visits.

Differences in demographic and medical characteristics between patients with and without missed follow-up appointment at 6 months after surgery arthroscopy were compared using the Wilcoxon rank-sum test for continuous variables and chi-square or Fisher exact tests for categorical variables. Similar tests were used to compare differences in demographic and medical characteristics between patients with and without missed follow-up appointment at 12 months after surgery arthroscopy. Logistic regression was used to examine associations of demographic and surgical characteristics with missed follow-up appointment at both 6 and 12 months after surgery arthroscopy. Any variable with a P-value <0.25 in initial analysis was a candidate for multiple logistic regression. Only variables that were significant at P < 0.05 were retained in the final model.

There were 449 patients included in the study, of which 106 (24%) involved an isolated rotator cuff repair and decompression. The median age for the included patients was 57 years (interquartile range [IQR], 51 to 62), and 248 (55%) were women. The median state ADI decile was 6 (IQR, 4 to 6), and the median national ADI percentile was 24% (IQR, 16% to 29%). Additional patient demographics are given in Table 1. The median follow-up at the time of final database access was 1515 days (IQR, 980 to 2,091). The prevalence for follow-up at 3, 6 weeks, 3, 6, and 12 months was 95%, 69%, 83%, 57%, and 40%, respectively. For the entire cohort, 234 patients (52%) were treated in the outpatient setting by another orthopaedic surgeon within the faculty practice at some time point after the shoulder arthroscopy, of which 8 (1.8%) underwent a subsequent arthroscopy of any joint within 12 months after the initial shoulder arthroscopy. Of the 268 patients who did not return for the 12-month follow-up, 187 (69.8%) had no additional visits with their operating shoulder surgeon postoperatively and were considered LTFU. However, 80 of these LTFU patients were subsequently seen by another member of the department for another orthopaedic issue. Overall, therefore, the prevalence for LTFU from the practice was 24%.

The patient demographics and surgical data for those who did and did not return for the 6-month follow-up are given in Table 2. Having commercial insurance, involvement in workers' compensation, being a former or non-smoker, and being employed as a physical laborer or office worker was associated with a reduced risk of missing the 6-month follow-up visit. Furthermore, missing a visit within the first 3 weeks (P < 0.001) and at 3 months (P < 0.001) was associated with an increased risk of missing the 6-month follow-up. Distance from home to clinic and implementation of an electronic patient contact system were not associated with an improved follow-up at 6 months (P = 0.128 and P = 0.189, respectively). Multivariable analysis demonstrated that having commercial insurance (odds ratio [OR], 0.38; 95% confidence interval [CI], 0.23 to 0.62; P < 0.001) or workers' compensation (OR, 0.15, 95% CI, 0.05 to 0.43; P < 0.001) and being more socioeconomically disadvantaged as measured by a higher state ADI (OR, 0.86; 95% CI, 0.77 to 0.97; P = 0.015) or higher national ADI (OR, 0.988; 95% CI, 0.978 to 0.998, P = 0.021) were independently associated with a decreased risk of missing the 6-month follow-up visit. The most disadvantaged quartile of patients by state ADI data (OR, 0.51; 95% CI, 0.28 to 0.96; P = 0.035) were more likely, than the least disadvantaged, to return for follow-up, although the association did not hold for national ADI quartiles. By contrast, missing a postoperative visit within 3 weeks (OR, 1.77; 95% CI, 1.14 to 2.74; P = 0.01) and missing the 3-month visit (OR, 8.55; 95% CI, 4.33 to 16.86; P < 0.001) were independently associated with a missed 6-month follow-up visit (Table 3).

The patient demographics and surgical data for those who did and did not return for the 12-month follow-up are given in Table 4. Female sex, involvement in workers' compensation, utilization of an electronic patient contact system, and increased number of preoperative visits were associated with a reduced risk of missing the 6-month follow-up visit. Furthermore, missing a visit at 3 months (P < 0.001) and at 6 months (P < 0.001) was associated with an increased risk of missing the 12-month follow-up. Marital status and self-identified race were not associated with follow-up at 12 months (P = 0.813 and P = 0.778, respectively). Multivariable analysis demonstrated that an increased number of preoperative visits (OR, 0.91; 95% CI, 0.84 to 0.99; P = 0.033) and utilization of an electronic patient contact syst