Evaluating the Use of PROMs in Paediatric Orthopaedic Registries

Patient-reported outcome measures (PROMs) provide structured information on the patient’s health experience and facilitate shared clinical decision-making. Registries that collect PROMs generate essential information about the clinical course and efficacy of interventions. Whilst PROMs are increasingly being used in adult orthopaedic registries, their use in paediatric orthopaedic registries is not well known. The purpose of this systematic review was to identify the frequency and scope of registries that collect PROMs in paediatric orthopaedic patient groups. In July 2023, six databases were systematically searched to identify studies that collected PROMs using a registry amongst patients aged under 18 years with orthopaedic diagnoses. Of 3190 identified articles, 128 unique registries were identified. Three were exclusively paediatric, 27 were majority paediatric, and the remainder included a minority of paediatric patients. One hundred and twenty-eight registries collected 72 different PROMs, and 58% of these PROMs were not validated for a paediatric population. The largest group of orthopaedic registries collected PROMs on knee ligament injuries (21%). There are few reported dedicated orthopaedic registries collecting PROMs in paediatric populations. The majority of PROMs collected amongst paediatric populations by orthopaedic registries are not validated for patients under the age of 18 years. The use of non-validated PROMs by registries greatly impedes their utility and impact. Dedicated orthopaedic registries collecting paediatric-validated PROMs are needed to increase health knowledge, improve decision-making between patients and healthcare providers, and optimise orthopaedic management.


Introduction
Patient-reported outcome measures (PROMs) are tools that are designed to assess a patient's perception of their health-related quality of life and their functional health status without interpretation from a medical professional [1,2].Self-assessment, by means of a questionnaire, is considered the best method of evaluating patient-based outcomes, as any influence from a clinician or investigator is removed [2].By assessing a patient's subjective health experience and the consequence of any intervention [2], PROMs are an essential tool to understand the impact a condition has on an individual's symptoms and disability [3].PROMs are vital to shared clinical decision-making and patient-centred care as they provide key information regarding the natural history of conditions and the efficacy of interventions that can assist all healthcare stakeholders (patients, healthcare professionals/providers, and policymakers) facing healthcare decisions [4].The broad utility and high importance of PROMs are reflected in their widespread adoption and standardised use amongst regulatory bodies, such as the US Food and Drug Administration and the European Medicines Agency, both of which mandate the use of PROMs to support labelling claims [5,6].The use of PROMs has increased substantially in the field of orthopaedics over the last 20 years as the evidence for their importance has grown [1].Since 2009, it has been mandatory to use PROMs to report outcomes for certain elective surgeries in the United Kingdom.The National Health Service publishes data from PROMs following orthopaedic surgical procedures to help drive improvements in surgical performance and service delivery [7].
Evidence of the increased use of PROMs is seen in the growing number of orthopaedic registries that have adopted PROMs [1].Registries were first established in the fields of arthroplasty and trauma to monitor implant survival [1].However, in recent decades, the utility of registries has been demonstrated by understanding patient characteristics, improving the timing and safety of intervention, and optimising public health decisionmaking [8].If registries are large enough and include an adequate follow-up, they can provide an ideal platform for clinical trials, reducing resources required for prospective data collection [9].Registry data can also be used to assist in answering questions that are not practical or ethical to address by randomised controlled trials [10].By tracking health outcomes over time, it is possible to identify the under-utilisation of evidencebased practices and areas for improvement [11].There is strong evidence that registry information can drive continuous improvements in patient outcomes and adherence to guideline-recommended care [10].Registries, however, cannot achieve these goals without the inclusion of PROMs [8].For example, in arthroplasty registries, the use of PROMs is now considered essential to determine a valid understanding of treatment success.Similarly, the improved survival rate in trauma registries has highlighted the need to collect PROMs to measure quality of life after injury [12].
Despite the importance of PROMs, there is little consistency in the use of PROMs in paediatric orthopaedics, and their use is infrequent compared to adult orthopaedics [2,13].Furthermore, where PROMs are used, they are commonly not validated for paediatric populations [13,14].If PROMs are not valid in the assessed population, they cannot be relied upon to measure the true impact of an intervention or inform healthcare decisions [14].The standardised use of validated PROMs in paediatric orthopaedic registries is an essential step towards improving clinical care in paediatric orthopaedics [13,15].Whilst PROMs orthopaedic registries are utilised in adult populations to improve the safety and efficacy of healthcare, in addition to strengthening communication and understanding between patients and healthcare providers, little is known about the use of PROMs in paediatric orthopaedic registries.
To ensure that PROM collection in paediatric orthopaedic registries is valid and useful in improving clinical understanding and care, it is crucial to identify gaps and weaknesses in the current state of PROM collection.It is vital to establish the current state of PROM collection by paediatric orthopaedic registries in order to highlight the most pressing issues and challenges facing this field of research and guide the future creation of registries.The aim of this systematic review is to achieve this goal by identifying the frequency and scope of registries that collect PROMs in paediatric orthopaedic patient groups and highlighting factors that need to be addressed to improve their utility.

Description of Studies and Risk of Bias
Of the 259 included studies, the majority were observational cohort studies, with the exception of 91 case-control studies.The style and purpose of the studies differed greatly, as seen in Tables 1-4.The risk of bias score for all studies, using the NOS for cohort or case control studies, is provided in the final column of Tables 1-4.All studies achieved scores of high quality (7)(8)(9), with the exception of five studies, which were rated as moderate

Type of PROMs
The registries used 72 different PROMs, including 24 generic, 8 hip pathology-specific, and 14 knee-pathology-specific (Table 5).Amongst these 72 PROMs, 42 (58%) did not include any paediatric validation, and 7 (10%) included validation limited to those 16 years and over.In the 3 exclusively paediatric registries, all PROMs used were validated for paediatric populations, and amongst the 27 majority paediatric registries, 61% of the PROMs used were validated for those under 18 years of age.Regarding PROM collection frequency, 21% of the registries collected PROMs as a one-off, and the remainder collected them at multiple time points.The three most common PROM collection time points were pre-surgery, one-year post-surgery, and two years post-surgery, however, there was great variation across all registries.

Registries
Overall, 128 unique registries that included patients under the age of 18 years in their reported data sets were identified.There were three registries that included exclusively paediatric patients (Table 1), 27 registries that included a majority (>50%) of paediatric patients (Table 2), 16 registries that included a minority (33-50%) of paediatric patients (Table 3), and 82 registries that included a small minority (<33%) of paediatric patients.(Table 4).There were 27 knee ligament registries, 21 arthroplasty registries, 21 spine registries, and 21 hip preservation registries (Table 6).The scope of registries ranged from single hospital-based to international, with 56% (n = 72) of all included registries limited to a single-hospital scope.We identified 21 regional registries, 25 national registries, and 10 international registries.(Figure 2).Of the 27 knee ligament registries that included patients under the age of 18 years, 16 were hospital-based registries, and 4 were national registries: the Danish, Swedish, Norwegian, and New Zealand Knee Ligament Registries [108,165,184,190].One registry was a majority paediatric hospital-based registry that used only PROMs validated for those under 18 years (Pediatric-International Knee Documentation Committee (Pedi-IKDC) and Children's Health Questionnaire(CHQ)) [29].The remaining 26 registries were minority paediatric but had notably larger proportions of patients aged under 18 years compared to the arthroplasty registries (Table 6).These registries used 23 PROMs, including 11 generic PROMs and 12 knee-specific PROMs.The two most used knee-specific PROMs were the Knee Injury and Osteoarthritis Outcome Score (KOOS), which is validated for those 16 years and over, and the International Knee Documentation Committee (IKDC), which is not validated for paediatrics.

Lower Limb Arthroplasty Registries
The lower limb arthroplasty registries included a small minority of paediatric patients, with the exception of one [28].Most were hip arthroplasty registries, of which two were national registries, with the majority being limited to a single-hospital scope [142,153].There were three that included hip, knee, and ankle arthroplasties in one registry [143,145] There were nine anatomy-specific and eight generic PROMs used by these registries (Table 5).The most commonly used were the Visual Analogue Scale (VAS), European Quality of Life-5 dimensions (EQ5D), and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), which were each used in four different registries.Of these, the WOMAC is not validated for paediatrics, the EQ5D is validated for those 16 years and over, and the VAS is validated for paediatric patients from the age of five years.

Spine Registries
There were 21 spine registries that included patients under the age of 18 years.Only 1 was exclusively paediatric [22], and a further 15 reported a majority of paediatric patients (Table 2).The most frequently used PROM was the Scoliosis Research Society Questionnaire (SRS) (various versions), which has been validated for the paediatric population from the age of 10 years.In both majority and minority paediatric registries, this PROM was occasionally used amongst participants younger than 10 years [22,84].Other PROMs used and validated for paediatric patients included the Early-Onset Questionnaire (EOSQ24) and the Caregiver Priorities Child Health Index of Life with Disabilities (CPCHILD) [55,61].Similar to the SRS, the Short Form 12 and 36 (SF12, SF36), the Body Image Disturbance Questionnaire (BIDQ), and the European Quality of Life 5 Dimensions 3 Levels (EQ5D3L) were all used in patients below the age of their paediatric validation range, and the Oswestry Disability Index (ODI) was used in spine registries despite not being validated for those under the age of 18 years [93, 94,227].

Hip Preservation Registries
We identified 21 hip preservation registries that included patients under the age of 18 years.A total of 2 of these had a majority of paediatric patients [30], and 18 were hospital-based.These 21 registries used 11 PROMs, including 8 hip-specific PROMs.Of these, only the Hip Outcome Score (HOS) was validated for patients under 18 years and utilised in 5 of the 21 hip preservation registries (Tables 1-5).

Discussion
This review highlights the paucity of PROM collection amongst paediatric patients by orthopaedic registries; specifically, only three dedicated paediatric registries collect PROMs in paediatric orthopaedic populations.There were an additional 125 orthopaedic registries that included both adults and paediatric patients, with 98 of these registries including a minority of individuals aged under 18 years.Of all studies reporting these registries, 98% were of high quality, with a low risk of bias.Registries that collect PROMs typically establish a structure for studies that avoids a number of risks associated with single studies, including bias in-patient selection, comparability of cohorts, prospective data collection, and duration of follow-up.Whilst these concerns are usually not an issue for a well-designed registry, the challenge of an adequate response rate, which was the NOQAS criterion most frequently not met by the studies in this review, can be a significant concern.
The importance of well-designed and well-maintained registries that minimise loss to follow-up has been widely established in adult populations [1].Such high-quality registry data have resulted in improved models of care in a number of health specialties.Some examples include accelerated ulcer healing time, attributed to the Swedish Ulcer Registry [345], and established causes of mortality associated with rheumatoid arthritis [346].Furthermore, diabetes registries have improved attendance at appointments and compliance with treatment regimens [347] and the Australian Breast Device Registry detected three devices with high complication rates, which were subsequently removed by the Therapeutic Goods Administration, resulting in reduced national revision rates [348].Likewise, in orthopaedics, data from the Australian Joint Replacement Registry identified high revision rates associated with the ASR™ Hip Resurfacing System, leading to a substantial reduction in their use and an overall reduction in hip and knee arthroplasty revisions since the registry has been in operation [349].The Victorian Orthopaedic Trauma Outcomes Registry identified key factors in demographics and injury management affecting return to work and mortality in those under 65 years who sustain a hip fracture [245,246].
The second largest proportion of registries identified in this review were arthroplasty registries that consistently use PROMs not validated for use in people aged under 18 years.Whilst the average age of patients undergoing arthroplasty was greater than 70 years in the early 1990s, in recent years, the average age has decreased, and future projections indicate that it will continue to do so [350].In light of the historically older age, it is not surprising that arthroplasty registries were not established with paediatrics in mind [350].However, given the documented increased frequency of paediatric arthroplasty [351][352][353], it is now essential that registries accommodate paediatric patients.The majority of the remaining orthopaedic registries identified in this review concern specific diagnostic groups such as knee ligament reconstruction, hip preservation procedures, spine surgery, and trauma.It is paramount that registries for these diagnostic groups collect validated PROMs for the age range of included children so that information gathered can be utilised to improve the clinical course of these conditions and gauge the efficacy of interventions [13].
One barrier to the inclusion of paediatric-validated PROMs in orthopaedic registries may be the limited number of appropriate PROMs available for specific diagnostic groups.Currently, the only hip-specific PROM with paediatric validation is the Hip Outcome Score, which is validated for those aged 13 years and over [305].A systematic review of hip PROMs used in older paediatric patients did not comment on whether the PROMs used were validated for the reported age group [354].Likewise, the lack of adequate PROMs is a significant challenge shared by rare disease diagnostic groups with orthopaedic involvement.The use of non-validated custom questionnaires by many of the rare disease registries highlights the inadequacy of existing validated PROMs for their purposes [21,101,102].A lack of validated PROMs significantly reduces the extent to which orthopaedic registries can capture relevant and valid information to ultimately improve healthcare efficacy and safety [13,355].
This review shows that when paediatric-validated PROMs are available, they are rarely used by orthopaedic registries that include paediatric patients [356,357].A challenge in using paediatric-validated PROMs in registries that include both adults and paediatric patients may be the increased burden of customising PROM delivery according to age [3].This was apparent in the knee ligament registries, which overwhelmingly used the KOOS [112,172] and/or the IKDC [119,358], and not the KOOS-child, validated from 16 years of age, or the Pedi-IKDC, which is validated and recommended for those under 18 years of age [315,359].Improved registry design to collect valid data from all patients that can be utilised to understand the natural history and surgical outcomes from childhood through to adulthood is required.The burden of integrating paeditric and adult versions of a PROM in the same registry can be overcome with digital platforms, such as research electronic data capture (REDCap) [360], which can automatically distribute age-appropriate validated PROMs.
Another possible reason for registries not using validated paediatric PROMs when available may be the challenge of comparing scores between paediatric and adult-version PROMs [3].This again can be overcome by using paediatric and adult versions of the same PROM that have published equivalency scores [359].By doing so, such registries would improve the understanding of orthopaedic conditions, and the impact of interventions as paediatric patients transition into adulthood.The integration of scores between two different PROMs remains a substantial challenge.Further research to establish the clinical and statistical relationship between the most appropriate paediatric and adult PROM will only be possible if appropriate validated PROMs are used in these registries.
The findings of this review point to two key actions that can be undertaken to improve PROM collection by orthopaedic registries.Firstly, for adult registries that include participants under the age of 18 years, accommodations must be made for these younger participants to ensure the data that are collected are valid and useful.Secondly, there is a need for further dedicated paediatric orthopaedic registries that collect PROMs in order to answer future questions concerning paediatric orthopaedic conditions and interventions.Such actions may be accelerated if policies are introduced by health services that require more uniform PROM collection amongst orthopaedic populations such as has been seen in arthroplasty registries [4].Furthermore, insistence on the use of validated PROMs by journals would result in registries no longer using non-validated tools.These changes have the potential to transform the scope and quality of paediatric orthopaedic research.Such improvements would increase the understanding of how orthopaedic conditions affect children and raise the standard of care provided to such children.
We acknowledge the limitations of this review.First, our search criteria included any registry that included patients under 18 years of age.This resulted in a large number of registries that included a very small proportion of paediatric patients, including a number of registries that included one or two 17-year-olds.However, we attempted to make this issue transparent by grouping the registries by the proportion of paediatric patients they included (Tables 1-4).Second, the exclusion of craniofacial orthopaedic diagnoses was undertaken due to a large overlap with dental medicine publications, as these were considered too far removed from the common understanding of paediatric orthopaedics.Further reviews examining the relevance of these articles may be indicated.Third, we acknowledge there may be registries in existence that collect validated PROMs in paediatric orthopaedic populations but have not yet published their findings and were, therefore, not included in this systematic review.

Conclusions
Currently, there are only three reported registries with publications that have been established to collect PROMs in paediatric orthopaedic patients, though many adult orthopaedic registries include the collection of PROMs in paediatric patients.Comparing this small number to the frequency of adult orthopaedic registries highlights the paucity of paediatric orthopaedic registries that collect PROMs.Given that these three registries report data collected since 2000, it is apparent that this is an area of clinical research that has been slow to change.The lack of systematic collection of validated PROMs in paediatric orthopaedics through registries means that the paediatric orthopaedic literature is largely dependent on clinician-reported outcomes and individual studies.This reduces the understanding of conditions and treatment impact from the perspective of the patient.As a result, the research findings may be limited by patient numbers and a narrower scope of investigated questions.In contrast, registries that collect PROMs provide essential information about the course of clinical conditions and interventions from the patient's perspective, ultimately promoting patient-centred care and shared decision-making.Therefore, if we are to better understand health conditions, assess interventions and improve the quality and safety of care in paediatric orthopaedics, registries must be established and must use validated PROMs in their target populations.An investment in infrastructure to support the collection of PROMs by registries in paediatric orthopaedics is needed from health service providers and policymakers.Such changes will allow health outcomes to be assessed in children and tracked as children grow into adults.

Figure 1 .
Figure 1.PRISMA flow chart of study selection.

Children 2023 ,
10, x FOR PEER REVIEW of 54

Figure 2 .
Figure 2. Scope of registries that include patients under the age of 18 years.
Key: # Study refers to more than one registry, = PROM validated for age range in study, = age range of study is outside validated range of PROM, ACL: anterior cruciate ligament, CC: case control study, CKRS: Cincinnati Knee Rating System, HOOS: Hip Disability and Osteoarthritis Outcome Score, HOS-ADL: Hip Outcome Score-Activities of Daily Living, HOS-SSS: Hip Outcome Score-Sport-Specific Subscale, IKDC: International Knee Documentation Committee, KOOS: Knee Injury and Osteoarthritis Outcome Score, K-SES: Knee Self-Efficacy Scale, MARS: Marx Activity Rating Scale, mHHS: Modified Harris Hip Score, NAHS: Non-Arthritic Hip Score, OC: observational cohort study, OHS: Oxford Hip Score, PAS: Physical Activity Scale, PD: published data, RAND-HSI: RAND Health Status Inventory, RTS: Return to Sport Questionnaire, SF12: Short Form-12, SF36: Short Form 36, UCLA: University of California Los Angeles Activity Scale, UK: United Kingdom, USA: United States of America, VAS: Visual Analogue Scale, WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index.

Table 4 .
Cont.Short Musculoskeletal Function Assessment, SRS: Scoliosis Research Society, SST: Simple Shoulder Test, SSV: Subjective Shoulder Value, TAA: total ankle arthroplasty, TESS: Toronto Extremity Salvage Score, THA: total hip arthroplasty, TKA: total knee arthroplasty, UCLA: University of California Los Angeles Activity Scale, UMWPAR: Unspecified Measure of Work, Physical Activity, and Restriction, USA: United States of America, UK: United Kingdom, VAS: Visual Analogue Scale, vNRS: Numerical Rating Scale, WBPQ: Web-Based Pain Questionnaire, WOMAC: Western Ontario and McMaster Universities Arthritis Index, WOSI: Western Ontario Shoulder Instability Index.
Key: # study refers to more than one registry, = PROM validated for age range in study, = age range of study is outside validated range of PROM, ACL anterior cruciate ligament, ACLR: anterior cruciate ligament reconstruction, ACL-RSI: Anterior Cruciate Ligament-Return to Sport after Injury Scale, ASES: American Shoulder and Elbow Surgeons Shoulder Score, bMHQ: Brief Manchester Hand Questionnaire, CC: case control study, CKRS: Cincinnati Knee Rating System, COPS: Core Questionnaire for the Assessment of Patient Satisfaction, EQ5D: EuroQol 5 Dimensions, EQ5D3L: EuroQol 5 Dimensions 3 Levels, HADS: Hospital Anxiety and Depression Scale, HOOS: Hip Disability and Osteoarthritis Outcome Score, HOS: Hip Outcome Score, HOS-ADL: Hip Outcome Score-Activities of Daily Living, IEQ: Injustice Experience Questionnaire, iHOT: International Hip Outcome Tool, IKDC: International Knee Documentation Committee, IPAQ: International Physical Activity Questionnaire, IPO-Q: International Pain Outcome-Questionnaire, KOOS: Knee Injury and Osteoarthritis Outcome Score, KOS-ADL: Knee Outcome Survey Activities of Daily Living Scale, K-SES: Knee Self-Efficacy Scale, KS-F: Knee Society-Function, LEFS: Lower Extremity Functional Scale, MARS: Marx Activity Rating Scale, mHHS: Modified Harris Hip Score, MODEMS-E: Musculoskeletal Outcomes Data Evaluation and Management System-Expectations, MoxFQ: Manchester Oxford Foot and Ankle Questionnaire, NAHS: Non-Arthritic Hip Score, NHS: National Health Service, NZ: New Zealand, OA: osteoarthritis, OC: observational cohort study, ODI: Oswestry Disability Index, OHS: Oxford Hip Score, OKS: Oxford Knee Score, PAO: Periacetabular Osteotomy, PCL: posterior cruciate ligament, PCLR: posterior cruciate ligament reconstruction, PD: published data, PG: Press Ganey, PROMIS ® : Patient Reported Outcomes Measurement Information System ® , PRWE: Patient-Rated Wrist Evaluation, PTSD: post-traumatic stress disorder, QWB: quality of well-being, RA: rheumatoid arthritis, RAND_HSI: RAND Health Status Inventory, RTW: return to work, SANE: Single Assessment Numeric Evaluation, SAS: Shoulder Activity Score, SEFAS: Self-Reported Foot Ankle Score, SF12: Short Form 12, SF36: Short Form 36, SMFA:

Table 5 .
PROMs used among paediatric patients in orthopaedic registries.

Table 6 .
Types of registries that include patients under the age of 18 years.Scope of registries that include patients under the age of 18 years.

Table 6 .
Types of registries that include patients under the age of 18 years.