Assessing Patient Recovery After Hip and Knee Replacement
Assessing Patient Recovery After Hip and Knee Replacement
All prospective longitudinal studies which reported change in the OHS or OKS from before to after surgery were screened for review. Studies which included adults with osteoarthritis of the hip or knee undergoing total hip or knee replacement were included regardless of the type of prosthesis, or rehabilitation protocol used. Studies which focused on hip resurfacing or unicompartmental knee replacement procedures were not included in the review. Only longitudinal studies that reported the mean OHS or OKS before and at 12 months after surgery were included. To reduce the amount of heterogeneity in the review, studies were excluded if the sample included patients undergoing revision procedures as this is a less effective procedure than primary surgery. We also excluded studies that used non-English versions of the OHS and OKS as this was considered a further source of heterogeneity. All potentially eligible studies were examined for the possibility of multiple-publication of data on a single patient sample. Where such studies were discovered only the original publication was included.
MEDLINE and the Web of Knowledge were searched for studies published in the period up to and including June 2011. In MEDLINE we reviewed all studies retrieved by searching for the text terms "Oxford hip", "Oxford knee" and "Oxford 12" in all fields. No time, language or study design restrictions were used. In the Web of Knowledge we searched for all studies which cited the original validation papers of the OHS and OKS. We used a conservative approach to title and abstract review and discarded only those studies which obviously did not satisfy our inclusion and exclusion criteria.
The PubMed search revealed a total of 94 citations that included the term "Oxford hip" and 129 that included the term "Oxford knee". The Web of Knowledge search identified 202 articles which had cited the original validation paper for the Oxford Hip Score and 199 articles which had cited the original paper for the Oxford Knee Score. Citations were managed using the Endnote software package. Following the elimination of duplicate papers, independent screening of the abstracts for these citations by two reviewers (JB and HB) identified 35 possibly relevant papers for the Oxford Hip Score and 31 possibly relevant papers for the Oxford Knee Score. Full text copies of these papers were obtained and independently reviewed for relevance by JB and HB. Papers which covered the wrong patient population or surgical procedures, presented insufficient data, were literature reviews or covered the same sample of patients as another paper were eliminated at this stage. Following discussion it was agreed that six papers for the Oxford Hip Score and nine papers for the Oxford Knee Score should be included in the final systematic review.
Two reviewers (HB and JB) independently extracted data on the OHS and OKS at baseline, 12-months and any subsequent time points from each included study. Any differences in the data extracted were resolved through discussion. For many studies it was necessary to convert the reported means from the original 12–60 scale (60=most severe) to the new 0–48 scale (0=most severe) which is now the accepted standard. The formula for converting scores was y = 60 - x where y is the desired score on the new scale and x is the score from the original scale. Change scores were extracted if reported and are of the same magnitude, irrespective of whether the 12–60 or 0–48 scoring system is used. If the change score was not reported it was derived by a simple subtraction of the reported pre-operative mean from the reported post-operative mean. Results for multiple patient groups are presented for comparative studies unless scores for all patient groups were presented in aggregate form somewhere in the paper.
The English PROMs Programme uses two important methods to minimise bias. First, all eligible patients are invited to take part in the Programme so that local hospital staff do not introduce selection bias. Second, all questionnaires are completed by the patient on their own so that interviewer bias on the part of local hospital staff or clinicians is avoided. To assess the extent to which the studies iin our systematic review had the same level of methodological quality as the English PROMs Programme, two authors (HB and JB) independently reviewed the methods used in the selection of study participants (to assess for potential selection bias) and the methods used in the administration of the OHS and OKS (to assess for potential interviewer bias).
To enable a comparison with data from the PROMs programme the Health and Social Care Information Centre in England was asked to provide the mean six-month improvement in the OHS and OKS for patients undergoing primary surgery. This request was necessary because although information about the PROMs programme is freely available on the internet it does not distinguish between patients undergoing primary and revision surgery. The data extract captured the primary part of the relevant Operating Procedure Codes (OPCS version 4.3) for total primary hip and knee replacement procedures and therefore matched the patient population covered by our systematic review.
The primary objective of this systematic review was to compare the post-operative improvements in the OHS and OKS observed at six months by the PROMs programme with the improvements at 12 months reported in the published literature. A direct statistical comparison was not possible because there was insufficient information in the published literature to derive a pooled estimate of change with confidence intervals at 12 months using a meta-analysis. Instead we provide a narrative synthesis which compares the clinical importance of change at six and 12 months. When comparing mean change scores we defined a minimally important difference (MID) as equal to or greater than one Standard Error of the Measurement (SEM). The SEM was chosen because it is relatively constant when measured in different samples of patients and therefore lends itself to comparison of results from different studies. The SEM has been estimated as 2.7 points for the OHS and 2.1 for the OKS. A second objective of the review was to compare the difference between 12-month outcome scores and outcome scores recorded at later time points.
Methods
All prospective longitudinal studies which reported change in the OHS or OKS from before to after surgery were screened for review. Studies which included adults with osteoarthritis of the hip or knee undergoing total hip or knee replacement were included regardless of the type of prosthesis, or rehabilitation protocol used. Studies which focused on hip resurfacing or unicompartmental knee replacement procedures were not included in the review. Only longitudinal studies that reported the mean OHS or OKS before and at 12 months after surgery were included. To reduce the amount of heterogeneity in the review, studies were excluded if the sample included patients undergoing revision procedures as this is a less effective procedure than primary surgery. We also excluded studies that used non-English versions of the OHS and OKS as this was considered a further source of heterogeneity. All potentially eligible studies were examined for the possibility of multiple-publication of data on a single patient sample. Where such studies were discovered only the original publication was included.
MEDLINE and the Web of Knowledge were searched for studies published in the period up to and including June 2011. In MEDLINE we reviewed all studies retrieved by searching for the text terms "Oxford hip", "Oxford knee" and "Oxford 12" in all fields. No time, language or study design restrictions were used. In the Web of Knowledge we searched for all studies which cited the original validation papers of the OHS and OKS. We used a conservative approach to title and abstract review and discarded only those studies which obviously did not satisfy our inclusion and exclusion criteria.
The PubMed search revealed a total of 94 citations that included the term "Oxford hip" and 129 that included the term "Oxford knee". The Web of Knowledge search identified 202 articles which had cited the original validation paper for the Oxford Hip Score and 199 articles which had cited the original paper for the Oxford Knee Score. Citations were managed using the Endnote software package. Following the elimination of duplicate papers, independent screening of the abstracts for these citations by two reviewers (JB and HB) identified 35 possibly relevant papers for the Oxford Hip Score and 31 possibly relevant papers for the Oxford Knee Score. Full text copies of these papers were obtained and independently reviewed for relevance by JB and HB. Papers which covered the wrong patient population or surgical procedures, presented insufficient data, were literature reviews or covered the same sample of patients as another paper were eliminated at this stage. Following discussion it was agreed that six papers for the Oxford Hip Score and nine papers for the Oxford Knee Score should be included in the final systematic review.
Two reviewers (HB and JB) independently extracted data on the OHS and OKS at baseline, 12-months and any subsequent time points from each included study. Any differences in the data extracted were resolved through discussion. For many studies it was necessary to convert the reported means from the original 12–60 scale (60=most severe) to the new 0–48 scale (0=most severe) which is now the accepted standard. The formula for converting scores was y = 60 - x where y is the desired score on the new scale and x is the score from the original scale. Change scores were extracted if reported and are of the same magnitude, irrespective of whether the 12–60 or 0–48 scoring system is used. If the change score was not reported it was derived by a simple subtraction of the reported pre-operative mean from the reported post-operative mean. Results for multiple patient groups are presented for comparative studies unless scores for all patient groups were presented in aggregate form somewhere in the paper.
The English PROMs Programme uses two important methods to minimise bias. First, all eligible patients are invited to take part in the Programme so that local hospital staff do not introduce selection bias. Second, all questionnaires are completed by the patient on their own so that interviewer bias on the part of local hospital staff or clinicians is avoided. To assess the extent to which the studies iin our systematic review had the same level of methodological quality as the English PROMs Programme, two authors (HB and JB) independently reviewed the methods used in the selection of study participants (to assess for potential selection bias) and the methods used in the administration of the OHS and OKS (to assess for potential interviewer bias).
To enable a comparison with data from the PROMs programme the Health and Social Care Information Centre in England was asked to provide the mean six-month improvement in the OHS and OKS for patients undergoing primary surgery. This request was necessary because although information about the PROMs programme is freely available on the internet it does not distinguish between patients undergoing primary and revision surgery. The data extract captured the primary part of the relevant Operating Procedure Codes (OPCS version 4.3) for total primary hip and knee replacement procedures and therefore matched the patient population covered by our systematic review.
The primary objective of this systematic review was to compare the post-operative improvements in the OHS and OKS observed at six months by the PROMs programme with the improvements at 12 months reported in the published literature. A direct statistical comparison was not possible because there was insufficient information in the published literature to derive a pooled estimate of change with confidence intervals at 12 months using a meta-analysis. Instead we provide a narrative synthesis which compares the clinical importance of change at six and 12 months. When comparing mean change scores we defined a minimally important difference (MID) as equal to or greater than one Standard Error of the Measurement (SEM). The SEM was chosen because it is relatively constant when measured in different samples of patients and therefore lends itself to comparison of results from different studies. The SEM has been estimated as 2.7 points for the OHS and 2.1 for the OKS. A second objective of the review was to compare the difference between 12-month outcome scores and outcome scores recorded at later time points.
Source...