but significant differences in more functional outcomes, such as wheezing.71 In this case, it seems straightforward for a clinician to choose the treatment option that results in better patient outcomes. But functional outcomes such as wheezing, or even health care utilization measures such as readmissions after NICU discharge, are more variably defined and reported than proximal morbidities such as BPD.72 Obtaining consensus on best treatment strategies based on patient-centered outcomes requires more effort in selecting and defining outcomes that matter to a broader range of stakeholders. In general, designing and powering studies that measure patient-centered outcomes in neonatal CER will continue to be a challenge. RCTs are often powered for a set of primary endpoints, such as death or BPD, and measure multiple secondary endpoints, such as respiratory symptoms. Interpreting results of multiple secondary outcomes is difficult when adjusting for multiple comparisons. Involving families explicitly in study design could provide additional insights into selection of outcomes from the multiple available options. In the United Kingdom, many health research funding bodies require patient involvement in study design. Consumer involvement has been reported in convincing funders of study relevance, developing clear and relevant study questions and outcomes, providing insights into patients’ views of trial logistics, advising on recruitment, and developing patient information materials.73,74 Neonatal research has noted differences in parent and clinician perspectives about neonatal treatments for extremely preterm infants and infants with trisomy 13 and 1875,76; differing perspectives are also likely present and insightful for decision making that involves less life-and-death situations. If the large sample size available for observational studies could facilitate evaluation of multiple outcomes, it could be easier to evaluate enough patient-centered outcomes to appeal to multiple stakeholders. Currently, many observational data sources lack discrete data on patient-centered outcomes besides mortality and length of stay. Measuring symptoms such as pain and fatigue rather than laboratory values or definitions (BPD) presents unique issues in a very young pediatric population because they require either clinician assessment or parent report, and there are no validated multidomain quality of life measures for infants under 1 month of age. Two-year neurodevelopmental assessments are often used as proxy measures for long-term functioning, although the correlation between 2-year assessments and later functioning is imperfect.77 There are many longitudinal studies of quality of life in survivors of preterm birth,78 but none have been used yet for CER. However, this could be an area for future development in neonatal patient-centered outcomes research. Nurses already record discrete pain scores, and therapists perform standardized motor assessments, which are part of the health record and could be abstracted for observational research.79 For infants older than 1 month of age there are validated tools for infants and young children, as well as for caregivers related to their child’s illness.80,81 In adult 840 Lagatta et al populations with wider use of patient-reported outcome measures, such as the National Institutes of Health-supported Patient-reported Outcomes Measurement Information System (PROMIS) measures, there are increasing efforts to embed patient-reported outcomes assessments into the electronic health record for clinical and research purposes.82,83 There is a need for more work on defining key patientcentered measures for infants in the NICU and in postdischarge follow-up to help focus efforts in neonatal CER. SUMMARY CER is a relatively new focus area that encompasses principles of clinical research that to some extent have been present in neonatal research for some time. Identifying best practices within the wide variation of NICU care can enable clinicians and quality improvement efforts to either standardize care across groups, or tailor efforts to specific patients that benefit from particular approaches. This increased emphasis on evidence to direct decision making ties in well with neonatology’s already strong efforts in quality improvement, systematic reviews, patient registries and multicenter efforts. Many data sources exist with the potential to increase CER efforts within neonatology. Future work is needed to define patient-centered outcomes to focus prospective clinical studies, and embed appropriate tools within observational data to facilitate analysis of patient-centered outcomes. REFERENCES 1. Sox HC, Greenfield S. Comparative effectiveness research: a report from the Institute of Medicine. Ann Intern Med 2009;151:203–5. 2. Luce BR, Drummond M, Jonsson B, et al. EBM, HTA, and CER: clearing the confusion. Milbank Q 2010;88:256–76. 3. Varadhan R, Segal JB, Boyd CM, et al. A framework for the analysis of heterogeneity of treatment effect in patient-centered outcomes research. J Clin Epidemiol 2013;66:818–25. 4. Wong J, Shah PS, Yoon EW, et al. Inotrope use among extremely preterm infants in Canadian neonatal intensive care units: variation and outcomes. Am J Perinatol 2014. [Epub ahead of print]. 5. Lagatta J, Clark R, Spitzer A. Clinical predictors and institutional variation in