To our knowledge, this is the first literature review that summarizes definitions of poor outcome after primary TKA presented in the literature. We found a total of 46 different definitions varying in nature and number of outcome domains involved, the type of response and the magnitude of change. A total of eight different dimensions were used in identified definitions of poor outcome: pain, function, physical functioning, health-related quality of life (HRQOL), patient satisfaction, anxiety, depression and patient global assessment. Patient satisfaction was used as single domain with a wide variation in wording of questions and answering categories. The absolute cut-off value was the most common type of response, with large variety in value and timing of follow-up. Our review stresses the need for an unambiguous definition for poor response after TKA to enable comparisons of the effectiveness of TKA among studies and among countries.
A remarkable finding of our review was that the majority of definitions used to describe poor outcome incorporate only one or two outcome domains. This finding does not seem to correspond with the conclusions by the OMERACT-OARSI initiative and the International Consortium for Health Outcomes Measurement (ICHOM). The OMERACT proposed a simplified set of responder criteria for (non-surgical) treatment of OA in clinical pharmaceutical trials. This set of responder criteria comprises relative and absolute changes in three domains: pain, physical functioning and global assessment (18,19). Also, the OMERACT TJR Working Group proposed a set of core domains (pain, function, patient satisfaction, revision, adverse events, and death) to evaluate joint replacement in randomized controlled trials (20). Parallel, ICHOM has developed a set of patient-centered outcome measures and case-mix factors for evaluating, comparing and improving the treatment (both surgical and non-surgical) of patients with hip and knee OA, focusing on outcomes that matter to patients (21). Pain, function, HRQOL and work status formed the core outcome domains, after a modified Delphi process (21). Corresponding to this standard set and these responder criteria it seems important to measure poor response to TKA within multiple domains to cover important key outcome domains to patients.
Some studies used the inverse of the OMERACT-OARSI responder criteria “nonresponse” as a definition of poor outcome (22–26). However, it is questionable whether “poor response” is the true opposite of “clinically meaningful response” as this definition implies that patients with smaller improvements will be part of the poor response group. The study by Mahler et al. showed a clear asymmetric magnitude of change, with a lower amount of change for patients who reported being worsened compared to the amount of improvement in patients who reported being improved (27). Therefore, in our opinion the amount of absolute or relative change in relevant domains is therefore an important aspect of definitions of poor outcome.
Patient dissatisfaction
Patient satisfaction was used as single domain with a wide variation in wording of questions and answering categories, most frequently measured by single item questions (non-validated instruments) (14). However, patient satisfaction is a multidimensional concept that may represent either satisfaction with outcome (e.g. knee function) of TKA or the process of care delivery, which all can be influenced by patients’ expectations (14,28). Halawi et al. explored subjective reasons for patient dissatisfaction after TJR and found different causes of patient dissatisfaction. The most common causes for dissatisfaction after TKA were persistent pain, functional limitation, surgical complication and reoperation, staff or quality of care issues and unmet expectations (28). It is likely that different factors influence the construct of patients satisfaction, and therefore it is important to determine the different determinants that contribute to patients satisfaction after TKA according to the perspective of patients and orthopaedic surgeons.
Different perspectives
The variety in definitions of poor outcome used could reflect different perspectives of different stakeholders. There are many studies reporting on the disagreement between the patient and physician in terms of their satisfaction with surgery (11,29,30). It is conceivable that physicians tend to focus on aspects of their clinical evaluation (e.g. stability, range of motion and alignment), while patients are more likely to focus on the functionality of the knee during daily life activity. Moreover, the view of physicians and patients on the desired magnitude of improvement after TKA is not always concordant, as poor correlations were found between physician-assessed and PROMs (13). Furthermore, most outcome measures have been developed according to the medical research perspective, which mainly address knee-specific measures like pain and function scores, and scarcely address mental functioning and consequences for social participation (31). So far, the choice for definitions to describe response or non-response after TKA has been dominated by non-comprehensive physician-based scoring systems and PROMs in quantitative research but the perspectives of patients and orthopaedic surgeons regarding the definition of poor response have been relatively neglected.
This study has some limitations, as our searches for relevant articles were systematic but the data extraction and writing were performed in a narrative manner. However, this inventory review does provide a complete overview of definitions of poor response after TKA that could be of interest to a large group of physicians and researchers involved in defining outcomes after TKA. Furthermore, only studies published in English language were included. For this reason, it cannot be ruled out that some studies were not identified (language bias).
In conclusion, this inventory review shows that many different heterogeneous definitions, incorporating several domains, for poor response to primary TKA are being used in the literature. Future research should focus on the perspectives and perceptions of orthopaedic surgeons and patients about constructs underlying poor response to TKA. Our findings stress the need for an consensus-based unambiguous definition of poor response to draw conclusions about the prevalence of poor-responders to TKA across hospitals and countries, and to identify patients at risk.