Much investigation into the outcomes of surgical correction of adolescent idiopathic scoliosis (AIS) has been undertaken to assess the ultimate goals of surgery, obtaining a balanced 3-dimensional correction of the deformity and achieving a solid arthrodesis. Increasingly more emphasis is being placed on the need to understand and evaluate patient-reported outcomes when assessing the efficacy of surgical intervention.
Although pain is typically not the primary factor for the patient with AIS to seek treatment or for surgeons to initiate operative management, it is a critical factor to analyze in understanding the efficacy and outcomes of surgical intervention. Follow-up into adulthood of patients treated with Harrington rods has shown that the majority of patients experience little to no back pain, at a rate approximately 10% greater than controls, but that the percentage increases with a more distal fusion level.1–3 Pain after surgery with more modern instrumentation techniques have been evaluated largely with the Scoliosis Research Society (SRS) questionnaires, which have a domain dedicated to assessing pain frequency, severity, medication usage, and impact on school attendance.4 Studies have shown that surgical correction in AIS significantly improves overall average pain domain scores at 2 years postoperatively compared with preoperative scores.5–8 Sanders et al9 identified an overall rate of 5% of patients with poor results at 2 years in the SRS-22 Pain domain. Landman et al6 found that at 2 years after surgery, 64% of patients reported mild to severe pain within the prior 6 months.
The current investigation was undertaken to further understand the prevalence of postoperative pain as reported by patients with AIS to the treating surgeon/clinical team within the 2-year postoperative period. Additionally, the impact that self-reported pain has on SRS-22 outcome instrument scores was evaluated. Lastly, the preoperative profile of patients who self-report postoperative pain was also evaluated using SRS-22 outcome tool scores.
MATERIALS AND METHODS
A prospectively enrolled multicenter AIS database was queried to identify patients for this study. The inclusion criteria for enrollment into the database registry includes: diagnosis of AIS with a coronal Cobb in the operative range for which operative management is recommended, male or female, and age 10 to 21 years at time of enrollment. All sites maintain institutional review board approval and signed informed consent is obtained on all patients prior to enrollment.
Patients with 2-year follow-up, preoperative SRS-22 scores, and 2-year SRS-22 scores were identified via a query of the larger database. Descriptive data, preoperative SRS-22 scores, 2-year SRS-22 scores, and specifically reported complication data were obtained. Pain, as reported by the patient to the treating surgeon/clinical team in follow-up subsequent to hospital discharge (first recorded visit at 6-week follow-up), is recorded by the sites as a potential complication in the database and was included in the query. Patients included in this series were identified as either reporting pain (pain group) or not (pain free group) to the surgeon/clinical team. The postoperative visit at which the pain was reported was also recorded. This was categorized into patients reporting postoperative pain (within 6 mo of surgery) and those reporting pain after the postoperative period (6 to 24 mo after surgery). The postoperative pain group was further stratified in an attempt to isolate pain that may be explained by another existing complication (explained pain) or pain reported by a patient in whom no other cause was identified (unexplained pain). Patients with unexplained pain after the initial 6-month postoperative period (pain reported 6 to 24 months after surgery) were the cohort of interest for analysis.
Descriptive data were calculated for the postoperative pain and postoperative pain free groups. Further descriptive information was calculated for the postoperative pain group (distribution of timing and explained/unexplained pain). Pre- and 2-year postoperative SRS-22 scores were then compared between the unexplained postoperative pain group and postoperative pain free group (pain reported 6–24 mo after surgery) using analysis of variance (P < 0.05). All analyses were conducted with SPSS version 12 (SPSS Inc., Chicago, IL).
Five hundred and eighty-four patients were identified. The average age at surgery was 14.7 ± 2 years (range, 10–21), with a majority of female patients (80%). The majority of the cohort (523) did not report pain within the 2-year postoperative period. Sixty-one (11%) patients did report pain at least once between the period after discharge from the hospital and the 2-year follow-up visit. Thirteen of the 61 (21%) were within the 6-month postoperative period. Of the remaining 48 patients reporting pain between 6 and 24 months postoperatively, 41 (7% of the total cohort) had no obvious cause for their pain. More than half of these patients (26/41) were referred for further treatment (physical therapy, referral to pain specialist, and further imaging).
The average age at surgery of the 41 patients with unexplained postoperative pain was 15.3 ± 2 compared with 14.6 ± 2 in the postoperative pain free group (P = 0.07). Both groups were comprised of a majority of female patients (postoperative pain group 83% female; postoperative pain free group 80% female, P = 0.619). The Lenke classification was similarly distributed across the 2 groups (P = 0.896, Table 1). There was also a similar distribution of anterior and posterior approaches between the 2 groups (postoperative pain group: 80% posterior, 20% anterior; postoperative pain free group: 87% posterior, 13% anterior; P = 0.21). The 2 groups had similar 2-year major coronal Cobb, coronal balance (C7–CSVL), and sagittal kyphosis (P > 0.05, Table 2). There were also no significant differences in terms of coronal decompensation, rib prominence on forward bend, and trunk shift (P > 0.05, Table 2). The postoperative pain group had significantly decreased 2-year SRS-22 scores in all domains compared with the postoperative pain free group, except for the domain of Function (Table 2, P < 0.05). The average Pain domain score for the patients who self-reported postoperative pain was significantly lower (lower score indicative of greater pain) than the postoperative pain free group (4.1 postoperative pain vs. 4.5 postoperative pain free, P < 0.001), and the average difference of 0.4 exceeds what has been deemed as the minimal clinically important difference (MCID) for the Pain domain (MCID = 0.2).10 The distribution of SRS-22 reported 2-year Pain domain scores for both groups are shown in Figure 1. A significantly greater proportion of patients without postoperative pain (29%) scored a 5 (highest possible) on the Pain domain (indicative of no pain experience and no pain medication use) than the patients with postoperative pain (12%, P = 0.024).
The 41 patients with unexplained postoperative pain were also found to have significantly decreased preoperative SRS-22 Pain domain scores (lower score indicative of greater pain) compared with the postoperative pain free group (P < 0.001, Table 3). This average difference was 0.4, which again is greater than the suggested MCID for this domain. Patients who had a SRS-22 Pain domain score less than 4 preoperatively were 2.6 times more likely to report unexplained postoperative pain (P = 0.003). There were no significant differences between the 2 groups for any other domains preoperatively (P > 0.05). The increase in SRS-22 Pain domains from pre- to postoperative was significant for the postoperative pain free group (P < 0.001); however, the increase for the patients with unexplained postoperative pain did not reach significance (P = 0.11).
The assessment of pain, particularly of chronic pain, is a challenging task. Pain is a personal, very subjective experience that is often heavily mediated by extraneous factors in an individual's life. Demographic characteristics of an individual have been found to influence how pain is reported and can be confounding factors in studies evaluating pain.11 Multiple methods of assessing and quantifying pain have been used in the assessment of postoperative pain in the patient with AIS. Thus the variability in reported prevalence of 2-year postoperative pain in this population should come as no surprise.6,9
Despite the differences in categorizing a “pain” outcome, our overall rate of 7% of patients with unexplained pain is similar to the overall prevalence of 5% of patients with “poor” SRS Pain domain score at 2 years reported by Sanders et al.9 Although our postoperative pain group was comprised of patients reporting pain as a symptom during their follow-up visit, and Sanders et al9 defined a poor outcome as being less than 2 standard deviations below the mean on the SRS questionnaire, the resultant percentage of patients was within 2% of each other. However, the previous study did find a higher rate of poor SRS pain scores in Lenke 3 and 6 patients, whereas this study found no significant variation in distribution of pain report across curve types. A study by Landman et al6 found that 64% of patients responded positively to the singular question assessing the presence and degree of pain they have experienced in the last 6 months, which is much greater than the 5% to 7% prevalence of reported pain found by Sanders et al9 and this study. It is unknown if this 64% is an accurate representation of patients with AIS experiencing 2-year postoperative pain or an overestimation of the pain experience as a result of using only one question intended to be quantified with other questions as part of a domain score.
Our group of patients who self-reported postoperative pain to their treating surgeon during their history on routine postoperative examination demonstrated significantly lower average scores on the Pain domain of the SRS-22 as than the patients within our cohort that did not report pain. This finding adds further external validation to the Pain domain of the SRS questionnaire. However, 12% of our postoperative pain group (by historical self-report) scored a 5 (indicative of no pain) on the SRS-22 Pain domain suggesting some discordance between the subjective history and the SRS-22 outcomes tool. In the postoperative pain free group, 14% scored below a 4 on the Pain domain of the SRS, indicating moderate to severe pain experience when none was reported to the clinician during routine examination. This concerning discrepancy may not necessarily be a fault of the SRS questionnaire. External factors have been found to play a significant role in self-report of pain. Sex discrepancy between the person reporting pain and the interviewer, environment, and time of day, all have been shown to influence pain reporting.12–15 It is possible that this 12% to 14% discrepancy between patient report to clinician and patient self-report on an outcome tool is the natural error inherent between these 2 types of assessments. The other possibility is the range in temporal variation between patient reporting of pain in this cohort (6–24 mo) and the completion of the 2-year SRS-22.
Not only did our group of patients with postoperative pain score significantly lower on the Pain domain of the SRS-22 questionnaire, but their scores in all other domains except for function were lower than patients who did not experience postoperative pain. In evaluating preoperative SRS scores, the postoperative pain group also demonstrated Pain domain scores indicative of greater pain preoperatively than the postoperative pain free group. These differences in multiple domains and in preoperative pain scores raise questions of whether; (1) the experience of postoperative pain is so severe that it affects other areas of personal experience, (2) patients who experience pain prior to surgery have expectations that are not met with the surgical correction of their scoliosis, or (3) psychological and/or personality differences exist between these groups and may possibly influence surgical outcomes. Evidence to support any and all 3 of these potential explanations exists in the literature.16–19 Future research into understanding this cohort of patients is essential in the era of evidence based medicine as the experience of pain has been shown to influence patient satisfaction and quality of life after treatment.18
This study is limited in that the review of the prospective database was retrospective in nature. We did not design prospective data collection with this study hypothesis in mind and with standardized questions related to the assessment of pain. The severity of pain experienced in the self-reported patients with postoperative pain was not assessed. The prevalence of pain reported here is likely influenced by various clinical examination protocols and documentation practiced by varying clinicians and centers. Pain reported by the patients is recorded by the sites as a potential complication; however, it is unclear whether the pain was thought by the surgeon to be directly related to their treatment or disease or an unrelated phenomenon. However, we think the rate of pain found here is translatable and representative of the population of patients with postoperative AIS who report pain to their surgeon/clinical team in the natural clinical setting. The multicenter nature of this study (with a variety of practices and geographical settings) makes the prevalence reported here more generalizable to the AIS surgical population than a single-center series.
The lack of a comparison group of age and sex matched controls limits us in truly concluding the clinical significance of the 7% incidence of postoperative pain in this cohort. A recent cross-sectional study of 1391 17-year olds found that 9.6% had a high probability of being diagnosed with low back pain and neck and/or shoulder pain.20 These patients had a low incidence of psychological comorbidities. They also identified a cluster of patients with moderate probability of low back pain, but high probability of psychological disorders (10.7%). The authors proposed that the mechanisms underlying the pain experience may be different in these different cohorts of patients. Unfortunately, previous medical history related to spinal disorders or treatment was not reported in this study. However, the results of that study suggest that it is possible that the 7% incidence of pain found in this study is in line with what would be identified in a control population of adolescents. Furthermore, these data indicate that the relationship with psychological comorbidities will exist for some, but not all of the patients reporting postoperative pain.
Other limitations are related to the differences in SRS-22 outcomes scores observed between the 2 groups. In this study, those differences were adequately powered to be significantly different, but the clinical relevance of these differences is yet to be determined. The postoperative pain group reported pain to the clinical team between 6 months and 2 years postoperatively, and analysis of the SRS-22 scores were those reported at the 2-year follow-up. Lastly, while we found no significant improvement in SRS pain scores within the group of patients who self-report pain postoperatively, this is likely due to the relatively few patients in that cohort representing a lack of statistical power.
In this series of patients, 7% of postoperative patients with AIS reported unexplained pain to their treating surgeon after the 6-month postoperative period. These patients had lower scores in the Pain domain of the SRS-22 outcomes instrument at 2 years of follow-up, thus further validating this measurement tool. This group of patients with postoperative pain also demonstrated decreased scores in other SRS domains compared with the postoperative pain free group. Interestingly, these patients also have lower preoperative pain scores than those without postoperative pain. Future study into the causes of preoperative and postoperative pain in this population is needed. Furthermore, understanding whether preoperative education and expectations targeted at this population would positively impact outcomes is warranted.
- Seven percent of patients with AIS in this series report unexplained pain to their treating surgeon/clinical team in the 6- to 24-month follow-up period.
- The patients who report unexplained pain have significantly lower 2-year SRS scores in all domains except function when than a cohort with no pain reported.
- These patients also had significantly lower Pain domain scores preoperatively than their pain free counterparts.
1. Bartie BJ, Lonstein JE, Winter RB. Long-term follow-up of adolescent idiopathic scoliosis patients who had Harrington instrumentation and fusion to the lower lumbar vertebrae: is low back pain a problem? Spine (Phila Pa 1976) 2009;34:E873–8.
2. Cochran T, Irstam L, Nachemson A. Long-term anatomic and functional changes in patients with adolescent idiopathic scoliosis treated by Harrington rod fusion. Spine (Phila Pa 1976) 1983;8:576–84.
3. Grouw AV, Nadel CI, Weierman RJ, et al. Long term follow-up of patients with idiopathic scoliosis treated surgically: a preliminary subjective study. Clin Orthop Relat Res 1976;117:197–201.
4. Haher TR, Gorup JM, Shin TM, et al. Results of the Scoliosis Research Society instrument for evaluation of surgical outcome in adolescent idiopathic scoliosis. A multicenter study of 244 patients. Spine (Phila Pa 1976) 1999;24:1435–40.
5. Carreon LY, Sanders JO, Diab M, et al. Patient satisfaction after surgical correction of adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 2011;36:965–8.
6. Landman Z, Oswald T, Sanders J, et al. Prevalence and predictors of pain in surgical treatment of adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 2011;36:825–9.
7. Merola AA, Haher TR, Brkaric M, et al. A multicenter study of the outcomes of the surgical treatment of adolescent idiopathic scoliosis using the Scoliosis Research Society (SRS) outcome instrument. Spine (Phila Pa 1976) 2002;27:2046–51.
8. Upasani VV, Caltoum C, Petcharaporn M, et al. Adolescent idiopathic scoliosis patients report increased pain at five years compared with two years after surgical treatment. Spine (Phila Pa 1976) 2008;33:1107–12.
9. Sanders JO, Carreon LY, Sucato DJ, et al. Preoperative and perioperative factors effect on adolescent idiopathic scoliosis surgical outcomes. Spine (Phila Pa 1976) 2010;35:1867–71.
10. Carreon LY, Sanders JO, Diab M, et al. The minimum clinically important difference in Scoliosis Research Society-22 Appearance, Activity, And Pain domains after surgical correction of adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 2010;35:2079–83.
11. Rosseland LA, Stubhaug A. Gender is a confounding factor in pain trials: women report more pain than men after arthroscopic surgery. Pain 2004;112:248–53.
12. Dworkin SF, Chen AC. Pain in clinical and laboratory contexts. J Dent Res 1982;61:772–4.
13. Folkard S, Glynn CJ, Lloyd JW. Diurnal variation and individual differences in the perception of intractable pain. J Psychosom Res 1976;20:289–301.
14. Jamison RN, Brown GK. Validation of hourly pain intensity profiles with chronic pain patients. Pain 1991;45:123–8.
15. Levine FM, De Simone LL. The effects of experimenter gender on pain report in male and female subjects. Pain 1991;44:69–72.
16. Adogwa O, Parker SL, Shau DN, et al. Preoperative Zung Depression Scale predicts outcome after revision lumbar surgery for adjacent segment disease, recurrent stenosis, and pseudarthrosis. Spine J 2012;12:179–85.
17. Breivik H, Borchgrevink PC, Allen SM, et al. Assessment of pain. Br J Anaesth 2008;101:17–24.
18. Dworkin RH, Jensen MP, Gould E, et al. Treatment satisfaction in osteoarthritis and chronic low back pain: the role of pain, physical and emotional functioning, sleep, and adverse events. J Pain 2011;12:416–24.
19. Kleiber C, Suwanraj M, Dolan LA, et al. Pain-sensitive temperament and postoperative pain. J Spec Pediatr Nurs 2007;12:149–58.
20. Beales DJ, Smith AJ, O'Sullivan PB, et al. Low back pain and comorbidity clusters at 17 years of age: a cross-sectional examination of health-related quality of life and specific low back pain impacts. J Adolesc Health 2012;50:509–16.