Article type
Year
Abstract
Background: chronic postsurgical pain (CPSP) after breast cancer surgery is associated with reduced quality of life, unemployment, and increased healthcare costs. There is considerable uncertainty regarding the prevalence of CPSP after breast cancer surgery, as published studies have reported rates that range from 10% to 90%.
Objectives: we conducted a systematic review of observational studies that reported rates of CPSP after breast cancer surgery to calculate a pooled rate across studies and explore potential sources of heterogeneity.
Methods: we searched MEDLINE, Embase, CINAHL, and PsycINFO from inception to October 2018 to identify observational studies with at least 100 patients that reported the rates of CPSP after breast cancer surgery. We performed random-effects meta-analysis with Freeman-Tukey transformation to calculate the overall prevalence of CPSP, and pooled pain intensity after converting all pain scales to a 10 cm visual analogue scale (VAS) for pain. We explored potential sources of heterogeneity for our pooled estimates.
Results: we included 188 observational studies with 300,906 patients, of which 165 studies (88%) were conducted in high-income countries, 23 (12%) in middle-income countries and none in low-income countries. The length of follow-up ranged from 3 months to over 10 years. Of the 188 eligible studies, 155 (82%) used representative samples and 129 studies (69%) used valid pain measures. Only 22 (12%) used a rigorous definition of CPSP (distinct from pain before surgery, associated with the site of surgery, and not explained by other reasons). Most studies (n =150) reported pain prevalence, with 22 (15%) using high threshold (more than mild pain) from pain instruments, 128 (85%) using low threshold (pain present or not). 126 studies reported pain intensity using 63 different instruments.
The pooled prevalence of CPSP was 34.2% (95% confidence interval (CI) 30.9% to 37.6%); pooled pain intensity on a 10 cm VAS was 3.0 (95% CI 2.7 to 3.4). The pooled prevalence of neuropathic pain, moderate-to-severe and severe pain was 31.2% (95% CI 23.7% to 39.2%), 20.7% (95% CI 17.7% to 23.9%) and 4.2% (95% CI 3.1% to 5.5%), respectively. (Figure 1)
We found significant subgroup effects between the use of 'any pain' versus localized pain to measure CPSP, low versus high threshold pain, and axillary lymph node dissection (ALND) versus sentinel node biopsy (Figure 2). Meta-regression did not show a significant association between CPSP and length of follow-up, proportion of loss to follow-up, proportion of breast-conserving surgery, radiotherapy, or chemotherapy.
Conclusions: CPSP after breast surgery affects between 24% to 44.6% of women, depending on how pain is measured and whether or not surgery involved ALND. Future studies should explore if nerve-sparing for axillary procedures could reduce CPSP after breast surgery. More studies in low- and middle-income countries are required.
Patient or healthcare consumer involvement: no.
Objectives: we conducted a systematic review of observational studies that reported rates of CPSP after breast cancer surgery to calculate a pooled rate across studies and explore potential sources of heterogeneity.
Methods: we searched MEDLINE, Embase, CINAHL, and PsycINFO from inception to October 2018 to identify observational studies with at least 100 patients that reported the rates of CPSP after breast cancer surgery. We performed random-effects meta-analysis with Freeman-Tukey transformation to calculate the overall prevalence of CPSP, and pooled pain intensity after converting all pain scales to a 10 cm visual analogue scale (VAS) for pain. We explored potential sources of heterogeneity for our pooled estimates.
Results: we included 188 observational studies with 300,906 patients, of which 165 studies (88%) were conducted in high-income countries, 23 (12%) in middle-income countries and none in low-income countries. The length of follow-up ranged from 3 months to over 10 years. Of the 188 eligible studies, 155 (82%) used representative samples and 129 studies (69%) used valid pain measures. Only 22 (12%) used a rigorous definition of CPSP (distinct from pain before surgery, associated with the site of surgery, and not explained by other reasons). Most studies (n =150) reported pain prevalence, with 22 (15%) using high threshold (more than mild pain) from pain instruments, 128 (85%) using low threshold (pain present or not). 126 studies reported pain intensity using 63 different instruments.
The pooled prevalence of CPSP was 34.2% (95% confidence interval (CI) 30.9% to 37.6%); pooled pain intensity on a 10 cm VAS was 3.0 (95% CI 2.7 to 3.4). The pooled prevalence of neuropathic pain, moderate-to-severe and severe pain was 31.2% (95% CI 23.7% to 39.2%), 20.7% (95% CI 17.7% to 23.9%) and 4.2% (95% CI 3.1% to 5.5%), respectively. (Figure 1)
We found significant subgroup effects between the use of 'any pain' versus localized pain to measure CPSP, low versus high threshold pain, and axillary lymph node dissection (ALND) versus sentinel node biopsy (Figure 2). Meta-regression did not show a significant association between CPSP and length of follow-up, proportion of loss to follow-up, proportion of breast-conserving surgery, radiotherapy, or chemotherapy.
Conclusions: CPSP after breast surgery affects between 24% to 44.6% of women, depending on how pain is measured and whether or not surgery involved ALND. Future studies should explore if nerve-sparing for axillary procedures could reduce CPSP after breast surgery. More studies in low- and middle-income countries are required.
Patient or healthcare consumer involvement: no.
PDF