17 Measuring Outcomes in Pediatric Spinal Deformity
Outcomes measurement is a fundamental component of evidence-based medicine, and it provides clinicians and patients with the information needed for shared and informed decision making. In pediatric spinal deformity care, outcome measurement has traditionally focused on the magnitude of the spinal curvature. However, in the last two decades health care systems have become more patient centered on an individual, hospital, healthcare service, and policy level. A focus on patient-centered outcome measures has implications for the daily practice of the spine surgeon as well as for patients and families, with the need for the surgeon and the family to become more actively involved in the management of the patient and the deformity. The wishes and needs of the patient require more measures than solely the radiographic ones. On a policy level, routinely measuring outcomes of care enables continuous evaluation and improvement. In the future, reimbursement may be based on value rather than on the quantity of care delivered. The transition from process measures to patient-based measures of outcome will empower surgeons and health care systems to focus on the patient′s health care experience.
A basic concept in outcome measurement is the distinction between outcome domains (e.g., pain, function, walking, appearance) and the instruments that are used to measure these outcome domains (e.g., Visual Analogue Scale, Oswestry Disability Index, Spinal Appearance Questionnaire). An outcome domain can be measured with different instruments; for example, physical functioning can be measured by a “stand up and go” test or by different questionnaires about functioning. Furthermore, outcome instruments can be categorized into clinician-based outcome instruments (e.g., functional tests) and patient-reported outcome measurements (PROMs), which are questionnaires that measure perceived outcomes.
The Use of Outcome Measures
Outcome measures can be used for the following purposes: (1) treatment planning and evaluation, (2) continuous quality improvement and value-based health care, and (3) clinical research. Ideally, the same measurement instruments and measurement moments could be used for all of these purposes. Unfortunately, this is not always feasible, as each purpose requires different qualities and characteristics of a measurement instrument. For the first two purposes, it is necessary to routinely measure the outcomes in clinical practice (i.e., every patient), whereas for most research purposes a (random) sample of patients may be adequate. Routine outcome measurement is challenging, but when implemented successfully it has many benefits. For example, research has shown that asking providers to measure and report outcomes improves their performance. 1 , 2 Understanding and comparing outcomes facilitates continuous learning and improvement of their own strategies through learning from best practices. To make fair comparisons between institutions (so-called benchmarking), outcome measures need to be corrected for preoperative status, and risk (case mix). Comparison between observed and expected outcomes may be useful in guiding system changes for improvement of outcomes. 3
Outcome Measurement for the Individual Patient′s Treatment Planning and Evaluation
For the individual patient, outcome measures can be used in the daily practice to gain more insight into their own functioning or quality of life. They can assist in the pretreatment workup (e.g., for risk assessment) and by measuring the change between pre- and posttreatment, they help in the evaluation of the treatment. By visualizing the results, they may improve communication and help manage expectations between the caregiver and the patient. In addition, outcome measures may also be used by patients for the evaluation of individual treatment trajectories (self-management and self-monitoring). Measuring outcome for the individual patient empowers the patient to participate in informed and shared decision making regarding health care and management of spinal disorders.
Outcome Measurement for Continuous Quality Improvement and Value-Based Health Care
Outcome measures are increasingly being used for continuous quality improvement. Research has shown that asking providers to measure and report outcomes alone already improves performance. 1 , 2 By providing a structure for testing of changes, quality improvement methods such as plan–do–study–act (PDSA) cycles have been used in an attempt to drive such improvements. This four-staged cyclic learning approach is aimed at the documentation of data at each stage of a treatment over time to understand natural variation in a system, increase awareness of factors that may influence outcome, and understand the impact of an intervention performance by evaluating treatment effectiveness. 4 This cycle of understanding and comparing outcome facilitates continuous learning and improvement of one′s own treatment strategies through learning from best practices.
Currently in many health care systems around the world, benchmarking of outcomes and process indicators is being performed, often by stakeholders such as regulators and payers (e.g., insurance companies). In the face of these fundamental transformations in health care, it is conceivable that in the near future multiple health care systems, policymakers, governments, and payers will adjust their reimbursements based on the value (outcome per unit cost) that is created for the patient. 5 A value-based health care economy prioritizes outcome of care over volume of care, and measuring outcomes is fundamental in creating a reimbursement system that optimizes providing cost-effective care.
Outcome Measurement for Research Purposes
In research settings, outcome measures are essential in evaluating treatment effectiveness and efficacy (randomized controlled trials [RCTs], cost-effectiveness studies) and in identifying prognostic and predictive factors that may enable providers to create predictive models.
Effectiveness and Efficacy Research
In the surgical management of spinal deformities, especially in pediatrics, RCTs are challenging to conduct and are often deemed unethical. As an alternative to RCTs, observational data from routinely collected outcomes are also useful for evaluating treatment strategies. 6 It is essential to bear in mind that treatment strategies are not randomly assigned, and therefore when comparing outcomes of treatments, it is essential to try to adjust the outcomes for confounding factors, such as preoperative risk factors.
Prognosis and Prediction Modeling
Prognostic factors are associated with clinical outcome in the absence of treatment. These variables are helpful for defining the natural history of disease, for identifying patients at risk of contacting the disease of interest, and for determining the progression of the disease.
Predictive factors are associated with clinical outcome following a particular treatment. These variables are helpful in identifying who will or will not benefit from a particular treatment. 7 Both prognostic and predictive factors rely on the outcome considered as an end point. Therefore, when performing research on prognostic and predictive factors for spinal deformity, it is essential to identify which outcome is being studied. For example, in cancer research, the 5-year survival rate is considered the outcome for estimating the prognosis of a particular form of cancer. For conditions such as spinal deformity, survival is not a relevant outcome. The (change in) quality of life is more relevant, which requires questionnaires evaluating the overall quality of life as an outcome for prognosis and prediction research.
Quality of Outcome Measurement/Clinimetric Properties
Measurement of patient-based health status, or patient value of a health condition, is challenging. The number of measurement instruments has increased over time, and the choice of which instrument to use has become more difficult. When choosing a measurement instrument, and this is true for questionnaires or patient-reported measures as well as for other more objective or clinician-derived measures, it is important to take into account the quality of the instrument. The study of Terwee et al 8 features the nine most important quality criteria of outcome measures: content validity, internal consistency, criterion validity, construct validity, reproducibility, longitudinal validity, responsiveness, floor and ceiling effects, and interpretability. Depending on the design, methods, and outcomes, each criterion can be rated as negative, positive, or indeterminate. These criteria have proven to be helpful in distinguishing between low- and high-quality measurement instruments.
Which Outcomes Are Important in Pediatric Spinal Deformities?
Outcome measures in pediatric spinal deformity, which encompass the whole cycle of care, measuring overall quality of life, functioning, and disability from a patient′s or caregivers perspective, will play an important role in defining appropriate care in future health care systems. The rapid growth in number and the abundant variety of measurement instruments can be confusing. Which outcome domain and instrument should be used, in light of the increasingly recognized shift toward a value-based health care from the patient′s perspective?
In the evaluation of scoliosis, generic instruments, such as the Short Form Health Survey (SF-36) and the EuroQol Group Health Questionnaire (EuroQol5D), can be used to assess health-related quality of life and cost evaluations, and to compare the health status with other diseases. Currently, the only condition-specific patient-reported outcome measurements is the Scoliosis Research Society (SRS)-22 questionnaire. The SRS-22 was introduced as a condition-specific outcome measurement for adolescent idiopathic scoliosis (AIS), and it measures five outcome domains: function, pain, self-image, mental health, and satisfaction. However, outcomes domains that are measured by any instruments can differ substantially in importance for children, adolescents, and adults with spinal deformity. For example, whereas patients with adult spinal deformity (ASD) seeking reconstructive surgery mostly want relief of symptoms (e.g., pain) and improvement of quality of life, patients with AIS mostly undergo reconstructive surgery to avoid curve progression and improve self-image, rather than relief of disabling symptoms. 9 In addition to general condition-specific questionnaires such as the SRS-22, highly specific patient-reported outcome (PRO) questionnaires to assess a single outcome domain have been developed, such as the Spinal Appearance Questionnaire (SAQ) and the Trunk Appearance Perception Scale (TAPS). 10 Recently, the development of a core outcome set for adolescent deformity surgery was initiated by the Core Outcome Set for Scoliosis (COSSCO) study group (see below). As another example, using the international consensus rounds by the Nordic Spinal Deformities Society (Sweden, Denmark, Finland, Norway, and the Netherlands), a core set of 14 outcome domains (11 patient reported and 3 clinician reported) has been identified for adolescent patients with spinal deformity undergoing reconstructive surgery based on the World Health Organization′s International Classification of Functioning and Disability (ICF). Clinician reported core outcome domains include “change in deformity,” “complications,” and “reoperation.” How the remaining 11 patient reported domains out of the 14 core domains are measured by widely used patient-reported measurement instruments is presented in Table 17.1 . Based on this analysis, a yet-to-be-defined respiratory questionnaire is needed to measure pulmonary fatigue and respiratory function.