As shown in Fig.1, a systematic search yielded 1930 articles. After removing 485 duplicates, 1445 papers were left for the title and abstract screening, and 1046 records that did not meet the inclusion criteria were further removed. Then, 39 full-text articles were reviewed, and 25 studies that were found not to meet the inclusion criteria were further excluded. Subsequently, 7 studies were rejected due to qualitative studies, the same authorship, incomplete data, or the population was not breast cancer. Finally, this systematic review included 7 articles, including 2 RCTs and 5 quasi-experimental studies. Two studies [19,20] with incomplete data were descriptive.
Study characteristics
The studies were published between 1994 to 2022 (Table 1). The studies were conducted in six country such as Japan [14], Korea [15,22], Brazil [16], Italy [17], the USA [18] and France [21]. The trials comprised 675 breast cancer patients, sample size ranged from 55 to 152. Effects were described according to outcome measures including body image (six studies) [15-18,21,22], self-esteem (five studies) [14,16-18,22], anxiety (six studies) [15-18,21,22], depression (five studies) [15-17,21,22], sexual function (two studies) [15,22], and quality of life (two studies) [14,17].
Participantscharacteristics
All studies focused on patients during treatment or recovery period. The studies by Ikeda M, 2020 (31%) [14], Panissi KC, 2021 (16.5%) [16], and Di Mattei VE, 2017 (25%) [17] had a small number of patients who had not undergone surgery. Besides, participants of the remaining study were postsurgical breast cancer patients, including patients who will be, are, or had completed adjuvant therapy.
Interventionscharacteristics
The types of interventions were mainly education program. Among them, one study (Kang D, 2022) [15] combined psychological intervention, while one study (Quintard B, 2008) [21] adopted only cosmetic treatment. The intervention components were mainly to teach makeup techniques and camouflage strategies, including wigs, eyebrows, etc. The number of interventions varied from 1 to 4 times, the duration varied from 1 day to 4 weeks, and the total duration varied from 2 to 16 hours. The intervention mode was mainly face-to-face group, while only one study (Quintard B, 2008) [21] adopted one-to-one treatment. Interventions were mainly performed by cosmetic specialists, with only two studies (Ikeda M, 2020; Kang D, 2022) [14,15] involving oncology nurse specialists.
Effect of appearance care on psychosocial outcomes
Body image
Body image was evaluated in six studies [15-18,21,22] involving 686 participants. Eight instruments were used to measure body image level: the Body-Image Questionnaire (BIC); the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Breast cancer module (EORTC QLQ-BR23); the Body Image Scale (BIS) (Hopwood et al., 2001); the Body Image States Scale (BISS); Body Cathexis Scale (BCS); Body Image Scale (Leon et al., 1979); Body Parts Satisfaction Scale (BPSS) and a 10-point scale.
Among the studies, five studies [16-18,21,22] assessed the immediate effect of appearance care on body image, only two studies [17,18] generated a significant result. Five studies were eligible for meta-analysis without showing an overall statistically significant effect on body image (SMD = 0.25, 95% CI -0.18 to 0.68, P = 0.26, I2 = 89%).
Five studies [15-17,21,22]assessed the short-term effects of appearance care on body image, three studies [15-17] generated a significant result. Five studies were eligible for meta-analysis without showing an overall statistically significant effect on body image (SMD = 0.03, 95% CI -0.35 to 0.40, P = 0.89, I2 = 85%) (Fig.2).
Self-esteem
Self-esteem was evaluated in 5 studies [14,16-18,22] involving 476 participants. All studies used the Rosenberg Self-Esteem Scale (RSES).
Among the studies, four studies [14,17,18,22] assessed the immediate effect of appearance care on self-esteem, only one study [18] generated a significant result. Four studies were eligible for meta-analysis showing an effect of appearance care on self-esteem (MD = 0.63, 95% CI 0.37 to 0.89, P< 0.00001, I2 = 0%).
Four studies [14,16,17,22] assessed the short-term effects of appearance care on self-esteem, two studies [16,17] generated a significant result. Four studies were eligible for meta-analysis without showing an overall statistically significant effect on self-esteem (MD = 0.08, 95% CI -0.20 to 0.36, P = 0.58, I2 = 90%) (Fig.3).
Anxiety
Anxiety was evaluated in six studies [15-18,21,22] involving 620 participants. Three instruments were used to measure anxiety level: the Hospital Anxiety and Depression Scale (HADS), the Beck Depression Inventory-II (BDI-II), and the Visual Analogue Scale(VAS).
Among the studies, five studies [16-18,21,22] assessed the immediate effect of appearance care on anxiety, four studies [16-18,22] generated a significant result. Four studies were eligible for meta-analysis showing a small effect of appearance care on anxiety (SMD = -0.46, 95% CI -0.60 to -0.31, P < 0.00001, I2 = 43%).
Five studies [15-17,21,22] assessed the short-term effect of appearance care on anxiety, four studies [15-17,22] generated a significant result. Five studies were eligible for meta-analysis showing a small effect of appearance care on anxiety (SMD = -0.42, 95% CI -0.54 to -0.34, P < 0.00001, I2 = 22%) (Supplementary Fig.1).
Depression
Depression was evaluated in five studies [15-17,21,22] involving 499 participants. Three instruments were used to measure anxiety level: the Hospital Anxiety and Depression Scale (HADS), the State-Trait Anxiety Inventory-Y Form (STAI-Y), and the Visual Analogue Scale(VAS).
Among the studies, four studies [16,17,21,22] assessed the immediate effect of appearance care on depression, two studies [17,22] generated a significant result. Four studies were eligible for meta-analysis showing a small effect of appearance care on depression (SMD = -0.41, 95% CI -0.62 to -0.19, P = 0.0001, I2 = 0%).
All five studies[15-17,21,22] assessed the short-term effect of appearance care on depression, four studies [15-17,22] generated a significant result. Five studies were eligible for meta-analysis showing a small effect of appearance care on depression (SMD = -0.41, 95% CI -0.55 to -0.26, P < 0.00001, I2 = 0%) (Supplementary Fig.2).
Sexual function
Sexual function was evaluated in 2 studies [15,22] involving 159 participants. Two instruments were used to measure sexual function: the Derogatis Sexual Functioning Inventory (DSFI) and EORTCQLQ-BR23.
Only one study [22] assessed the immediate effect of appearance care on sexual function, showing a significant result. While both two studies [15,22] assessed the short-term effect of appearance care on sexual function, all generated a significant result. Two studies were eligible for meta-analysis showing a moderate effect of appearance care on sexual function (SMD = 0.50, 95% CI 0.18 to 0.81, P = 0.002, I2 = 0%) (Supplementary Fig.3).
Quality of life
Quality of life was evaluated in 2 studies [14,17] involving 143 participants. Three instruments were used to measure quality of life level: Functional Assessment of Cancer Therapy-General (FACT-G) and EORTC QLQ-C30 and Skindex-16.
Among the studies, two studies [14,17] assessed both the immediate and short-term effect of appearance care on quality of life. Two tools of one study [14] generated a significant result on immediate effect, and Skindex-16 of one study [14] generated a significant result on short-term effect. Two studies were eligible for meta-analysis without showing an overall statistically immediate (SMD = 0.82, 95% CI -0.14 to 1.78, P = 0.10, I2 = 95%) and short-term (SMD = -0.08, 95% CI -1.33 to 1.18, P = 0.91, I2 = 97%) significant effect on quality of life (Supplementary Fig.4).
Other outcomes
Some studies assessed the effect of appearance care on happines [18], attractiveness [18], social support [22] and coping style [21,22], the results showed appearance care had a significant short-term effect on happiness and attractiveness, while the immediate and short-term effects of social support and coping style were not significant.
Sensitivity analysis
The included studies were deleted one by one for sensitivity analysis to evaluate the impact of a single study on the combined effect size. The results did not change substantially. That was to say, the meta-analysis results were relatively robust.
Publication bias
Given that no pooled analysis combined more than 10 studies, publication bias was not assessed [26].
Critical appraisal result and quality of evidence
All studies scored at least 50% on the JBI checklist and were therefore included in this systematic review. Among the RCTs, the study (Kang D, 2022) blinding of intervention practitioners and outcome assessors could not be achieved, and other items were scored as “yes”. The study (Quintard B, 2008) had five items (1-2, 4-6) that were all scored as “unclear”. Quasi-experimental studies (Sharon L, 1994 and Park HY, 2015) all applicable items were scored “yes”, three pre-and post-test studies (Panissi KC, 2021; Ikeda M, 2020; Di Mattei VE, 2017) failed to achieve to set up control group (iterm 4), all remaining applicable items were scored “yes”. The quality assessment results for the RCTs and the quasi-experimental were presented in Table 2 and Table 3.