What does the SRS-22 outcome measure tell us about spinal deformity surgery for Adolescent Idiopathic Scoliosis in the UK?

Abstract

Introduction

The purpose of this study was to analyse SRS-22 outcomes measures recorded on the British Spine Registry (BSR) for adolescent idiopathic scoliosis (AIS) surgery in the UK.

Methods

All cases having completed an SRS-22 outcome score and labelled with a diagnosis code of ‘AIS’ on the BSR were analysed. The SRS-22 score for primary cases was analysed by both individual domains and as a total score over time following surgery.

Results

A total of 3,860 cases were labelled as AIS recorded from 3,481 individuals. For primary cases, surgery improved the SRS-22 scores in every domain and as a total score, and this was maintained over time. There was no significant change in the scores recorded between 1 and 2 years of follow up apart from in function (and thus total score) for primary cases.

Conclusions

Surgery for AIS in the UK improves quality of life assessed using SRS-22. Mandatory follow up to 2 years postoperatively adds little information not already known at 1 year. We recommend that the Best Practice Tariff incorporates the collection of outcomes data as this is likely to reduce missing data.

Introduction

The British Spine Registry (BSR) was launched in 2012 as a method of recording the diagnoses, procedures and outcomes of spinal surgery in the UK in a systematic fashion. Consent is given at the time of data collection for that information to be used in research in an anonymised fashion at a later date. Features were also included to facilitate multicentre clinical trials. Within the BSR, different types of spinal pathology are classified using ‘pathways’, with scoliosis surgery found in the deformity pathway. Commissioning guidelines in England made the use of the BSR mandatory in 2013, but this proved difficult to enforce and a financial incentive scheme, known as the Best Practice Tariff (BPT) was introduced in 2019. Although the BPT mandates the recording of the surgical procedure only, it was also hoped that this would increase the recording of patient reported outcomes measures (PROMS).

As part of the BSR, outcomes are measured at fixed time points following surgery, using the outcome tool most suited to the pathology. In the deformity pathway, for anyone aged more than 10 years of age, the outcome measure used is the Scoliosis Research Society (SRS)-22 questionnaire.¹ The SRS-22 questionnaire comprises 22 questions, covering the domains of function, pain, mental health and self image. The format of the SRS-22 is that, for each domain and the total score, the result is presented out of a total score of 5, with 5 being the best possible score and 0 being the worst possible score. In the BSR, the SRS-22 is collected pre-operatively, at 6 weeks and 6 months postoperatively, and then at 1, 2, 5, 7.5 and 10 years. The SRS-22 questionnaire was designed and has been validated for use in adolescent idiopathic scoliosis (AIS).² The SRS-22 has not been validated in other conditions that are associated with the development of a scoliosis in an adolescent age group.

This paper uses the SRS-22 outcome questionnaire data that has been entered in the BSR to report on the outcomes of AIS surgery in the UK, and to analyse how the SRS-22 outcomes data can inform what surgeons should tell their patients to expect from surgery.

Methods

With the permission of the British Association of Spine Surgeons (BASS) and the British Scoliosis Society (BSS), all entries with SRS-22 outcome data from the BSR were made available for review, anonymised before release.

These data were then subdivided using the diagnosis code of AIS. The data were also assessed as primary cases (first and only procedure) and revision cases (more than one procedure). Box and whisker plots were used to examine the distribution of the data over the defined time points. A Kruskal–Wallis test was used to identify whether there was a statistical difference between the preoperative and all postoperative data across all time points. A Kruskal–Wallis test was also used to examine whether there was a significant difference in outcomes across longer term follow up (more than 1 year) from surgery. Statistical significance was defined in advance as p<0.05. All analysis was performed using R.³

Results

There were 16,439 SRS-22 outcome scores from 14,378 individual patients on the BSR when the data was examined in May 2020. There were 3,860 scores from 3,481 individuals with the diagnostic code AIS. Other diagnoses were also recorded and this is shown in Table 1. The distribution of data by age is seen in Table 2, and by the number of primary versus revision cases in Table 3. Of note, the numbers shown for each diagnostic category are greater than the total number of SRS-22 scores recorded, suggesting that multiple diagnoses were entered for 557 patients.

Table 1 .

The number of SRS-22 outcome scores recorded on the BSR by diagnosis

Diagnosis	n
AIS	3,860
Syndromic	557
Degenerative	402
Cerebral palsy	345
Juvenile idiopathic scoliosis	468
Adult idiopathic scoliosis	183
Congenital	283
Duchenne muscular dystrophy	36
Mucopolysaccharidosis	6
Meningomyelocoele	36
Neurofibromatosis	64
Other myopathy	69
Secondary	168
Spinal muscular atrophy	45
Syndromic	557
No diagnosis recorded	9,917
Total	16,996

Table 2 .

The number of SRS-22 outcome scores recorded on the BSR by age

Age range (years)	n
<10	853
10–18	4,461
>18	1,805
No age recorded	9,320
Total	16,439

Table 3 .

The number of SRS-22 outcome scores recorded on the BSR by number of procedures

Number of procedures	n
1	8,677
2	1,281
3	353
4 or more	427
No data recorded	5,701
Total	16,439

For primary procedures, Figures 1–5 show composite box and whisker plots for the separate SRS-22 domains and the total score. There were missing data given that data were not collected from every individual at every time point, and the number of responses for each time point demonstrates the amount of missing follow-up data, as shown in Table 4. In terms of percentages completing the SRS-22 at each time point for the primary group, 1,912 of 3,481 possibles (55%) had preoperative scores recorded. Of those completing a preoperative score, 64% recorded a 12-month score and 45% at score at 2 years (equivalent to 25% of the original overall total). The box plot demonstrates the median value as the heavy line in the box, with the box representing the interquartile range (IQR). The whiskers represent 1.5 times the IQR, and any points outside the whiskers are more than this from the median.

Figure 1 .

A composite box and whisker plot for the domain of function for primary surgeries for those with a diagnosis of adolescent idiopathic scoliosis

Figure 2 .

A composite box and whisker plot for the domain of pain for primary surgeries for those with a diagnosis of adolescent idiopathic scoliosis

Figure 3 .

A composite box and whisker plot for the domain of mental health for primary surgeries for those with a diagnosis of adolescent idiopathic scoliosis

Figure 4 .

A composite box and whisker plot for the domain of self image for primary surgeries for those with a diagnosis of adolescent idiopathic scoliosis

Figure 5 .

A composite box and whisker plot for the total score for primary surgeries for those with a diagnosis of adolescent idiopathic scoliosis

Table 4 .

The number of responses for each time point (n is the total number of entries on the BSR matching the criteria of a diagnosis of AIS for primary or revision cases)

	n with diagnosis of AIS	Pre-op score	6 weeks score	6 months score	12 months score	2 years score	5 years score	7.5 years score	10 years score
Primary	3,481 (100%)	1,912 (55%)	1,367 (39%)	1,290 (37%)	1,215 (35%)	865 (25%)	70 (2%)	25 (<1%)	29 (<1%)
Revision	379 (100%)	178 (45%)	96 (25%)	74 (20%)	50 (13%)	27 (7%)	1 (<1%)	0 (0%)	0 (0%)

Statistical analysis showed that there were significant differences in all domains and the total score (p<0.001) across the time points of data collection, from preoperative to 10-year follow up for primary procedures in the AIS group For the primary AIS group, statistically significant differences were seen only between the score at 1 year and the score at greater than 1 year in the function domain (p<0.001) and total score (p=0.034).

Discussion

Surgery for AIS is undertaken to prevent curve progression and improve the quality of life for the individual concerned.^4,5 The SRS-22 questionnaire was designed to assess the features deemed important to those with scoliosis and has domains that measure function, pain, mental health and self image, which then total for an overall score.¹ The SRS-22 questionnaire has been used extensively to record the outcomes of surgery for spinal deformity (a search using the PubMed search engine with the term ‘SRS-22’ returns 475 results).

The BSR acts to standardise collection of data on spinal procedures performed in the UK. The SRS-22 questionnaire is the main outcome measure used for AIS in the BSR. As the recommended data collection tool in the UK, the BSR acts as a source of information that can be used to assess the activity and outcomes for scoliosis surgery across the country.

This work demonstrates that, for primary AIS procedures recorded on the BSR, surgery changes the SRS-22 score. As might be expected, function and pain are reduced at the 6-week timepoint and this would be compatible with the early effects of surgery. However, by 6 months and then again at 12 months, pain and function has improved to that seen preoperatively. The self image domain scores more highly at 6 weeks postoperatively as compared with preoperatively, and this is maintained at all other time points. Mental health scores do increase over time from surgery, but the change preoperatively to postoperatively is slower, suggesting that, while surgery alters the shape (and thus the self image) of the body, there is less of an effect on the other components of mental health associated with a diagnosis of scoliosis. This would be in agreement with the work of Gallant et al, who reviewed the literature noting the associations of AIS with body image, eating disorders and mood.⁶

It is of interest to note that there is statistical significance only in the domain of function in the years that follow the first year of follow up. This may be related to the change in function generally seen between the adolescent teenage years and that of adulthood and may not be a true reflection of worsening function that may have been caused by surgery for AIS. Based on the data presented here, mandatory follow up to 2 years may not be useful to be able to measure the effects of scoliosis surgery with the SRS-22 questionnaire for the immediate postoperative period. However, there is not enough data here to comment further on how the SRS-22 scores might change over a longer term period (>2 years) after surgery.

The data from the BSR raises some other observations of note. There are a number of diagnoses recorded where the SRS-22 questionnaire has been used to record outcomes where the condition is not AIS. There are some where more than one diagnosis has been recorded for the same individual and some where the diagnosis and age are not matched appropriately.

There are also a number of cases in the AIS cohort that do not have any outcomes data. This highlights the potential downsides of the use of registry data, in that there may well be missing data and a number of data errors that cannot be addressed by a researcher who only has access to anonymised data from multiple sites. It also potentially highlights the lack of a suitable tool for the measurement of outcomes in spinal deformity surgery in the adolescent age group outside the condition of AIS, such that the SRS-22 questionnaire is thus used by physicians for lack of any suitable alternative.

It is recognised that the SRS-22 questionnaire has its faults with both floor and ceiling effects,⁷ and this may well lead to an underappreciation of the changes that occur in function, pain, mental health and self image following scoliosis surgery. The recognised mental health and self esteem issues that are seen in the AIS cohort may well be better assessed with more specific tools, in conjunction with the SRS-22 questionnaire.⁸

Of equal concern is the attrition rate seen with the capture of the data over time. As noted previously, in the primary surgery group, 1,912 of 3,481 (55%) had preoperative scores. Only 64% of those completing the preoperative score recorded a 12-month score, and this drops again to 45% at 2 years (25% of the overall total). For the collection of national PROMS to be of use and interest, there needs to be a much greater data capture rate at all time points. Without a more complete data set, any conclusions drawn will always be open to challenge. Making the collection of outcome data part of the BPT, in a similar way to that of the arthroplasty National Joint Registry,⁹ would improve this. Other things that could be considered to help improve the capture of follow-up data could include dedicated kiosks in the outpatient clinic, with support and guidance for those that require help alongside administrative support to follow up on missing data.

Given these caveats however, the SRS-22 questionnaire data from the BSR for AIS shows that, across the UK, scoliosis surgery improves the scores, both as the domains and as a total score. Using the SRS-22 score as an indicator for the quality of life of the individual, scoliosis surgery performed in the UK improves the quality of life of those who undergo it and this is maintained over the longer term. This is in line with what has been previously reported in the literature where SRS-22 questionnaire scores are seen to improve following surgery and have been maintained over time.^10–12

The strengths of this work are the large numbers of scores that are measured at standardised time points relative to the surgery, along with the national picture given by a national system used by the surgical community. The weaknesses of this work are the inability to add radiographic measures or clinical context to the outcome scores.

Conclusion

Deformity surgery for AIS in the UK, as measured by the SRS-22, improves quality of life after surgery that is maintained through to long-term follow up.