In this study, we investigated the incidence of CPSP after VATS in 2,348 patients. Our study showed that the incidence of CPSP was 43.99%. Most CPSP were mild and bearable, and the incidence of moderate-severe CPSP were 14.71%. The results of multivariate logistic regression analysis showed that age < 65 years, female, education level less than junior school, preoperative pain, consumption of rescue analgesia postoperative, consumption of sedative hypnotic postoperative, subcutaneous emphysema of chest wall postoperative, and history of postoperative wound infection were the risk factors of CPSP.
Previously published meta-analysis reported the incidence of CPSP after thoracotomy was 57%, but data about the risk of developing CPSP after VATS were not sufficient to summarize [3, 12]. Morten et al  did a randomized controlled trial to compare postoperative quality of life and pain between thoracotomy and VATS and found although VATS was associated with less postoperative pain for the first year after surgery, the incidence of CPSP after VATS was still as high as 30 to 50%. In Bayman et al’s  prospective observational study, they found no significant difference in the incidence of CPSP between thoracotomy and VATS (47% vs 33%). The high risk of developing CPSP after VATS may be explained by peripheral and central sensitization caused by local trauma, inflammation, or nerve injury . As indicated, the prevalence of CPSP reported in the present study was comparable to them. In addition to the high prevalence of CPSP, we were surprised to find that the majority of patients deal with CPSP in a negative way, only 15.23% of patients with moderate to severe pain sought active analgesic therapies (medicine or consultation to doctors), most of the remaining patients got relief from CPSP by resting or reducing activity. Overall, our results showed that although VATS significantly reduced the iatrogenic injury caused by surgical procedures, the incidence of CPSP after VATS remained high and most patients had a negative attitude towards CPSP management .
In terms of demographics data, age < 65 years, female and education level less than junior school were independent risk factors for CPSP. These results were consistent with previously published, known risk factors for CPSP in a variety of surgical procedures, including thoracic surgery [13–15]. A recent meta claimed that patients who developed CPSP after thoracic surgery were tend to be younger than patients without CPSP (mean difference 2.12 years, 95% CI -3.56 to -0.68) and female were found to be at slightly higher risk of CPSP (RR 1.15, 95%CI 1.02–1.30) . The relationship between age and CPSP can be explained by two points. First, young patients are biologically more sensitive to low-intensity noxious stimuli and may having a more heightened central nervous system responsiveness . Second, from a physiological perspective, older adults are more conservative in pain perception and reporting than younger adults and are more reluctant to report pain when it dose occurs . The association between sex and CPSP can also be explained from these two aspects. From biologically, differences of sex hormone levels, pain-related receptor activity such as N-methyl-D-aspartic acid receptor or P2X3 receptor, µ/κ subtype splits in the endogenous analgesic system, and brain structure and function between men and women are related to the mechanism of sex differences in pain perception . And from psychologically, female are more self-conscious and more likely to report pain to others . Association between education level and CPSP has also been proved in several studies, but the specific mechanism is still unclear, pain catastrophizing may be a mediating factor between them . Anyway, the combination of these demographic factors suggests a higher risk of developing CPSP. Recognition of these predictors, although unmodifiable, can help clinicians identify high-risk groups during preoperative evaluation and tailor an individualized pain treatment regimen .
As for perioperative pain-related parameters, our findings revealed that preoperative pain, the consumption of rescue analgesics and sedative hypnotic after surgery were independent predictors for CPSP. The association between preoperative pain and CPSP has been reported in several clinical trials. In a prospective study, Bayman et al  found those patients who reported higher preoperative pain scores at rest or upon coughing tended to have a higher likelihood of reporting chronic pain related to thoracic surgery at 6 months. Montes et al  also reported that preoperative pain at the surgical site and other sites was a risk predictor for the development of postoperative chronic pain in hernia repair, hysterectomy, and thoracotomy, which was confirmed by another risk index for CPSP . The mechanism between preoperative pain and CPSP remains elusive but cumulative evidence has demonstrated that sensitization of the peripheral and central nervous system, which related to the alterations of peripheral nociceptors sensitivity and function of the pain descending inhibitory system may be the possible explanations between them . Since preoperative pain could explain part of the interindividual variance in pain sensitivity, it may sensitize patients to new painful stimuli .
Postoperative use of rescue analgesics is an important indicator of postoperative pain intensity, especially when pain scores are performed only once or twice a day, the need for rescue analgesics helps in revealing the true level of postoperative pain intensity . Limited by the characteristics of the retrospective study, we were unable to collect a detailed information about postoperative acute pain score. Therefore, postoperative consumption of rescue analgesics and sedative hypnotic in hospital were investigated to reflect the existence of postoperative acute pain [2, 25]. Consistent with the thoracic surgery and other postsurgical chronic pain conditions, we reported that postoperative consumption of rescue analgesics and sedative hypnotic in hospital (which indicates a higher severity of acute pain) was associated with a greater risk of developing CPSP . Acute postoperative pain represents actual or potential tissue injury and motivates a response that removes the organism from such noxious stimuli . The more severe the postoperative acute pain, the more severe the tissue injury, and the less adequate the pain control, which may induce peripheral sensitization and neuroplastic changes that involves altered pain processing . Although our results did not show that nerve block reduces the risk of CPSP, the latest meta-analysis did report a meaningful change in the incidence of CPSP by controlling postoperative acute pain through regional anesthesia . Discrepancy between results may be related to differences in sample size, nerve block technique, and the type of study design. Therefore, the attempts to better manage postoperative acute pain are of great clinical significance in CPSP prevention.
For surgery related parameters, the history of postoperative wound infection and postoperative subcutaneous emphysema of chest wall were independently associated with the occurrence of CPSP.
Postoperative wound infection may be accompanied by severe, persistent, and recurrent inflammation . According to our results, there was no statistically significant difference in the contents of white blood cells, lymphocytes, neutrophils and CRP between the CPSP group and the non-CPSP group after surgery, but this may be related to the fact that we only routinely rechecked relevant biochemical indicators on the first day postoperatively. After discharge, some patients would suffer from out-of-hospital wound infection due to improper nursing or pleural effusion. However, due to the defects of retrospective study, we could not obtain relevant data comprehensively.
In addition, we identified a statistically significant relationship between postoperative subcutaneous emphysema of chest wall and the incidence of CPSP. We diagnosed patients with subcutaneous emphysema of chest wall based on the results of chest radiography (chest computed tomography or chest X-ray) during their postoperative hospitalization. No related literature has been found to prove that subcutaneous emphysema of chest wall is directly related to CPSP. The common causes of subcutaneous emphysema after surgery are leakage of lung tissue and inadequate thoracic drainage . When a patient develops subcutaneous emphysema, the surgeon will extend the indwelling time of the drainage tube or even re-place the thoracic drainage tube after ruling out other causes. The prolonged indwelling time and replacement of thoracic drainage tube may be the possible cause of the increased incidence of CPSP . Since it is the first time an association between subcutaneous emphysema and chronic pain has been found, stronger evidence is needed and the detailed mechanism of the association needs to be further explored. However, this finding reminds us of the importance of patency of postoperative thoracic drainage.
Our study has strengths, including the large sample size and comprehensive range of risk factors collection. However, the results should still be interpreted cautiously for several reasons. First, although we collected relevant data as comprehensive as possible, some potential predictors of CPSP remain missed. Information such as psychological status, the detailed postoperative acute pain score, and prolonged chest tube drainage should be included in future prospective study. And we speculated that the addition of these data may increase the area under the receiver operating characteristic curve. Second, since the incidence of CPSP decreased over time , the reported incidence of CPSP of our study may be still underestimated because of the long follow-up time span. Third, for the data collected from in a single institution, the results may be influenced by the selection bias.