Accuracy of post-operative recall by degenerative cervical myelopathy patients using the modified Japanese Orthopaedic Association scale

Abstract

Background and aim: The modified Japanese Orthopaedic Association (mJOA) scale is one of the primary measures of neurological function used on patients with degenerative cervical myelopathy (DCM). Contrary to some reports, the mJOA is not based on patient-reported outcomes as it is an assessment conducted by physicians, allied health professionals, or trained staff. To date, the accuracy of post-operative recall by DCM patients of their pre-operative neurological function, as assessed by the mJOA scale, has not been examined. This study, therefore, aimed to evaluate recall accuracy in DCM patients using the mJOA scale.

Methods: This study analyzed recall capacity of DCM patients who had undergone anterior cervical discectomy and fusion by a single surgeon at a large academic spine center between February 2012 and August 2017. Patient recall of neurological function pre-surgery was assessed at 3, 12, and 24 months post-surgery using the mJOA scale. Actual mJOA scores were also determined at each follow-up. Recall error (RE) was defined as the difference between recalled mJOA score at each post-operative visit and the actual baseline score. Age, gender, surgical segments, hospital length of stay, actual mJOA scores at follow-up, and actual rate of improvement in mJOA score were analyzed as predictors of recall accuracy. Descriptive statistics were collected to profile the characteristics of patients enrolled in the study cohort. All statistical computing and graphing were performed with R software and generalized estimating equation (GEE) model fitting was done using geepack package.

Results: A total of 105 patients (56.2% of males and 43.8% of females) were enrolled in the study. The median ± SD (range) age at the pre-surgical baseline measurement was 50 ± 8 (25 - 78) years. The recalled mJOA scores at the three follow-up time points were lower than the actual mJOA scores. The recall accuracy gradually decreased over time. Estimated coefficients showed that all variables in the GEE model except for surgical fusion segments were significant (P < 0.05). The pre-operative actual baseline mJOA score was inversely associated with RE. An increasing actual mJOA score over time had a significant positive influence on RE. Greater RE was found in males compared to females. Unexpectedly, age was inversely associated with RE.

Conclusions: The RE increases with the time interval between pre-surgical measurement and post-surgical follow-up and is more prominent in male DCMs patients following upper spine surgery.

Relevance for patients: It is necessary to select post-operative patients who need to pay attention according to the three factors of post-operative time, gender, and age, that is, patients with large RE should be given early or timely psychological counseling and treatment concerns, so as to reduce the occurrence of potential medical disputes and improve the level of medical safety.

Keywords: anterior cervical discectomy; degenerative cervical myelopathy; intervertebral fusion; neurological function; patient-reported outcomes; recall error.

Figure 1. Recalled and actual modified Japanese Orthopaedic Association (mJOA) scores at the three visits. JOA_3m_Re, JOA_12m_Re, and JOA_24m_Re signify recalled mJOA scores at 3, 12, and 24 months, respectively. JOA_3m_Ac, JOA_12m_Ac, and JOA_24m_Ac signify actual mJOA scores at 3, 12, and 24 months, respectively.

Figure 2. Scatter plot of recalled modified Japanese Orthopaedic Association (mJOA) scores at 3 months versus actual mJOA scores.

Figure 3. Scatter plot of recalled modified Japanese Orthopaedic Association (mJOA) scores at 12 months versus actual mJOA scores.

Figure 4. Scatter plot of recalled modified Japanese Orthopaedic Association (mJOA) scores at 24 months versus actual mJOA scores.

Figure 5. Recall error of the pre-operative actual baseline modified Japanese Orthopaedic Association score at each visit.

Figure 6. Scatter plot of recall error versus actual modified Japanese Orthopaedic Association score change stratified by gender.