Young or Old Age and Non-White Race Are Associated With Poor Patient-Reported Outcome Measure Response Compliance After Orthopaedic Surgery

Purpose To investigate orthopaedic patient compliance with patient-reported outcome measures (PROMs) and identify factors that improve response rates. Methods Our search strategy comprised a combination of key words and database-specific subject headings for the concepts of orthopaedic surgical procedures, compliance, and PROMs from several research databases from inception to October 11, 2022. Duplicates were removed. A total of 97 studies were included. A table was created for the remaining articles to be appraised and analyzed. The collected data included study characteristics, follow-up/compliance rate, factors that increase/decrease compliance, and type of PROM. Follow-up/compliance rate was determined to be any reported response rate. The range and average used for analysis was based on the highest or lowest number reported in the specific article. Results The range of compliance reported was 11.3% to 100%. The overall response rate was 68.6%. The average baseline (preoperative/previsit) response rate was 76.6%. Most studies (77%) had greater than 50% compliance. Intervention/reminder of any type (most commonly phone call or mail) resulted in improved compliance from 44.6% to 70.6%. Young and elderly non-White male patients had the lowest compliance rate. When directly compared, phone call (71.5%) resulted in a greater compliance rate than electronic-based (53.2%) or paper-based (57.6%) surveys. Conclusions The response rates for PROMs vary across the orthopaedic literature. Patient-specific factors, such as age (young or old) and race (non-White), may contribute to poor PROM response rate. Reminders and interventions significantly improve PROM response rates. Clinical Relevance PROMs are important tools in many aspects of medicine. The data generated from these tools not only provide information about individual patient outcomes but also make hypothesis-driven comparisons possible. Understanding the factors that affect patient compliance with PROMs is vital to our accurate understanding of patient outcomes and the overall advancement of medical care.

measurement, enabling hypothesis-driven comparison.Without PROMs, subjective data are highly heterogeneous, making comparison difficult. 1everal studies suggest incorporation of PROMs can improve patientephysician communication and patient outcomes. 1,2As U.S. health care costs increase, many services are under increased costcutting scrutiny.This has led to a rapid shift in reimbursement model from traditional volumedriven fee-for-service to value-based payment models. 3At the core of this shift is value analysis through PROMs. 4,5PROMs provide another measure to determine cost-effectiveness in health care.For this reason, clinical use of PROMs continues to increase at a rapid pace. 6s an objective measurement tool, it is essential for PROMs to have adequate responsiveness, validity, and reliability.7][8] Realworld compliance is multifactorial.Theoretically, variables including specific PROM used, method of admission, clinic staffing, and more may have significant effects on individual study compliance. 1,9Optimization of these variables is a common struggle experienced when incorporating PROMs into practice with no consensus on most important factors to consider. 5Due to this inconsistency, general compliance with PROMs in the field of orthopaedics is unknown.There is a paucity of information in the literature evaluating overall compliance regarding PROMs in the field of orthopaedics.Knowing PROM compliance rates is valuable in understanding potential for sampling bias, important factors of consideration in future clinical implementation, policy change, and study design.The purposes of this systematic review are to investigate orthopaedic patient compliance with PROMs and identify factors that improve response rates.Our hypothesis was that compliance to PROM would be poor but could be improved with the use of certain interventions.

Methods
This systematic review was conducted and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. 10

Information Sources and Search Strategy
Our search included MEDLINE, Embase, Cochrane Central Register of Controlled Trials (all via Ovid), Web of Science Core Collection, and SPORTDiscus via EBSCOhost from each database's inception until October 11, 2022.The search strategy comprised a combination of key words and database-specific subject headings for the concepts of orthopaedic surgical procedures, compliance, and PROMs.In order to capture the largest possible queue of articles, the only exclusion was non-English studies to avoid issues involving English translation.Some examples of key words include a combination of compliance or variations of the word (compliant, comply, complies, etc), PROM (PRO, PROM), specific PROMs (Patient-Reported Outcomes Measurement Information System, 12-Item Short Form Health Survey, Knee injury and Osteoarthritis Outcome Score, Western Ontario and McMaster Universities Arthritis Index, etc.) and orthopaedic surgery (ortho, orthopedic, arthroscopy, arthroplasty).

Selection and Data-Collection Process
After completion of the query, duplicates were removed using EndNote X9 (Clarivate Analytics, Philadelphia, PA).The identified articles were uploaded to Rayyan (Doha, Qatar) for screening.Initial screening of titles and abstracts for relevance was conducted by 2 independent reviewers (B.S.K., N.E.A.).Each reviewer was blinded to the results of the other to prevent any selection bias.Any discrepancies during the screening or extraction process were resolved by consensus agreement between the reviewers (B.S.K., N.E.A.) and the primary author (B.J.L.).Two separate rounds of review processes were performed.The first review was broader, including any study pertaining to orthopaedic surgery and PROMs.The second review was narrower, including only articles that specifically mentioned PROM compliance.Full texts of the remaining articles were obtained and assessed for eligibility by the same 2 independent reviewers in addition to the primary author.

Data Items
The information gathered from the systematic review was compiled into a table.The information included study characteristics, follow-up/compliance rate, factors that increase/decrease compliance, and type of PROM.Follow-up/compliance rate was determined to be any reported response rate.If different modalities were used in the study, those were included in the table.The range and average used for analysis was based on the highest or lowest number reported in the specific article.

Study Selection
The initial search yielded 4,035 citations.After removal of duplicates, 2,328 citations remained.After the first, broader screening, 1,500 citations remained.On the second, narrower screening, 97 were included (Table 1).  A flodiagram of the screening process is included in Figure 1.

Overall Compliance
All 97 studies reported PROM response in either the postoperative/postvisit setting or did not specify.The mean response rate overall was 68.6% (range 11.3%-100%).The median response rate was 73%.In total, 77% (75/97) of studies had greater than 50% compliance.

Results by Study Type
In total, 5.2% (5/97) of publications were randomized controlled trials (RCTs).][13][14] The 4 studies aimed to identify what factors improved response rate either compared with a control or to different modalities.The mean response rate among the RCTs was 54.8% (range 44%-71%, median 48%).
One RCT directly compared response rate based on different collection methods: phone call, e-mail, or mail. 15The overall response rate for the study was 48%.Phone calls yielded the greatest response rate of 64% versus 42% for e-mail and 42% for mail.In total, 94.8% (92/97) of citations were nonrandomized observational studies.The mean response rate among these studies was 69.4% (range 11.3%-100%).The median response rate was 75%.

Intervention
Intervention/reminder of any type (most commonly phone call or mail) resulted in improved compliance from 44.6% to 70.6%.Reminder types included phone call, mail, e-mail, text message, or some combination of

Discussion
The most important finding of this systematic review is that although a variety of factors can affect compliance with PROMS after orthopaedic surgery, reminders and other interventions can improve response rates.All 97 studies included in this systematic review reported PROM response rate in the postoperative setting.The average response rate across these studies was 53.6% (range 11.3%-100%).In addition to PROMs in the postoperative setting, it is crucial to obtain PROMs in the preoperative setting.Doing so establishes a baseline score for objective comparison to determine whether a surgical intervention was successful.Ideally, the rate of compliance in the postoperative setting should be similar or improved as compared with compliance in the preoperative setting.
Of the 97 studies that reported PROM compliance in the postoperative period, only 15% reported PROM response in the preoperative setting.The average response rate across these studies was 76.6% (range 7.4%-100%).When further examining the rate of PROM response in the postoperative setting for these 15 studies, the average response rate was 71% (range 40.6%-94.2%).Overall, the average PROM response rate in the postoperative setting for all included studies was 68.6% (range 11.3%-100%).
The compliance rates in PROMs poses several issues when evaluating the validity of an orthopaedic study.One particular concern is the introduction of response bias when patients are lost to follow-up.This could be attributed to a spectrum of reasons.One reason being these patients may experience worse outcomes in pain and function that discourage them from continued follow-up.In fact, 4 of the evaluated studies cited lower patient satisfaction as one of the reasons for decreased rates of PROM compliance.Socioeconomic and demographic factors may also play a role, as a number of the evaluated studies cited male sex, older age, non-White race, lower education, and lower income or unemployed backgrounds as risk factors for A commonly used method to increase PROM compliance is the use of reminders.In a study by Polk et al. 39 that observed PROM responsiveness in the Danish Shoulder Arthroplasty Registry, it was reported that the rate of response at the 1-year mark for followup was 65% before the use of a reminder.They then used mail-only and call/mail reminders to initial nonresponders, and subsequently observed response rates of 80% and 82% respectively.
PROM compliance also may depend on the mode of communication in which it is presented to patients.PROMs may be obtained with the use of surveys delivered via electronic or noneelectronic-based methods.This can include phone calls, mail or paper surveys, e-mail surveys, or SMS (ie, Short Message/ Messaging) responses.Overall, an intervention of any type demonstrated improvement in response rate from an average of 44.6% to 70.6% across all studies that used an intervention.Upon further analysis across 8 studies that used phone call-, electronic-, and mailbased interventions, phone call demonstrated the greatest compliance rate (71.5%) as compared with paper (57.6%) or electronic (53.2%).In a study by Schwartzenberger et al. 15 that implemented an RCT comparing phone, e-mail, and mail, they observed similar results, with telephone PROM collection having the greatest rate of compliance (64%) as compared with e-mail or mail (42% each).This may demonstrate the impact of personalized follow-up on compliance.Patients may feel more inclined to fill out a PROM survey when they are being directly asked.
Another consideration is that PROM surveys often contain medical jargon that is unfamiliar to patients, or patients may be unsure as to what particular PROM survey items are asking.Phone calls may help to address this potential issue and lead to an increase in compliance.This concept of personalized follow-up was further reinforced in one particular study by Tariq et al., 12 which used a last resort method of a personalized surgeon letter to individuals who did not initially respond to any interventions for follow up.They observed a 20% response rate in the intervention group as compared with 1.4% response rate in the control group that did not receive this letter.
We believe that this systematic review has strengths that may help to inform future orthopaedic research.We identified various patient-specific factors that may improve or reduce PROM compliance.In addition, this study was able to identify different means of intervention that could potentially lead to improved rates of compliance in PROMs collection.
It is important that orthopaedic researchers are aware of the potential impact that patient demographics may have on PROMs compliance.As reported within our study, male sex, extremes of age, and non-White race were cited as the most-common patient demographics associated with poor compliance rate.Early identification of these patients in the preoperative setting may be prudent, as focusing on these populations may generate different strategies that can be implemented to improve compliance within these groups moving forward.For example, in the younger population, it may be beneficial to obtain PROMs via SMS.As we move forward in a digital world in which the upcoming generations are being introduced to devices and internet access at a younger age, the use of electronic-based PROM surveys may soon become the norm.
Along these lines, orthopaedic researchers also should be aware of different interventions that may improve PROMs compliance.Patients can invariably be lost to follow-up for various reasons that may exist outside of a controlled research setting.As observed across many studies included in our review, phone calls, e-mails, and mail surveys represent successful methods that can lead to greater PROM response rates.

Limitations
There are several limitations that should be considered.The initial review process was conducted with 2 independent reviewers with 2 rounds of the screening process.Although this study design allowed for greater discretion of the proposed inclusion criteria, it is still possible that several studies may have been excluded unknowingly.In addition, several studies that cleared the initial screening process were ultimately not included in the final analysis due to unclear descriptions of patient characteristics or response rates.The vast majority of studies included for analysis were observational cohort studies, either prospective or retrospective, thus demonstrating only Level II or III evidence.Only 5 of the 97 total studies were randomized controlled trials demonstrating Level I evidence.It is also important to note that while the scope of this review was broad across general orthopaedic research, this also led to a heterogeneity of study designs that made it difficult to assess differences between studies.Some studies used broad PROMs such as EQ-5D or Patient-Reported Outcomes Measurement Information System computer adaptive testing, whereas other studies reported subspecialty specific PROMs such as Boston Carpal Tunnel Questionnaire or the Oxford Hip and Knee Score.It is difficult to discern whether PROMs response rates may vary depending on the type of PROM that is used.

Conclusions
The response rates for PROMs vary across the orthopaedic literature.Patient-specific factors, such as age (young or old) and race (non-White), may contribute to poor PROM response rate.Reminders and interventions significantly improve PROM response rates.

Fig 1 .
Fig 1. Flow diagram of search query.

Fig 2 .
Fig 2. Graph of response rate by survey type from the articles that were directly compared.

Fig 3 .
Fig 3. Graph of patient-specific factors cited as contributing to poor compliance.(BMI, body mass index.)

Table 1 .
Literature Review of the 97 Included Citations