As a result of searching electronic databases, 1371 articles were obtained, of which 1221 remained after removing duplicates. In the next step, after screening based on the title, abstract, and full text based on the inclusion and exclusion criteria, 18 studies [13-30] were selected for meta-analysis (Fig. 1), of which 17 studies [13-18, 20-30] reported the prevalence of depression and 16 studies [13-19, 22-30] reported the prevalence of anxiety in women with ovarian cancer. Of 18 included studies, 6 were conducted in Australia [13, 15, 20, 25, 26, 30], 2 in the USA [16, 17], 2 in the United Kingdom [18, 19], 2 in China [22, 23], and one in each of the following countries: Norway [14], Czech Republic [21], Thailand [24], the Netherlands [27], Italia [28], and Poland[29]. The average age range of women in these studies were from 36.5 to 62.1 years old (Table 1).
Prevalence of depression in women with ovarian cancer
The lowest depression prevalence belonged to the study of Hodgkinson K et al. [13] with a prevalence of 2% (95% CI: 0% - 4%), and the highest depression prevalence belonged to the Slovacek L et al. study [21] with a prevalence of 83% (95% CI: 70% - 97%). A total of 3085 women with different degrees of ovarian cancer were studied in these articles, of whom 611 were diagnosed with depression based on various depression measurement tools. After combining the results of these studies, the pooled estimate of depression prevalence was 27% (95% CI: 14% - 41%; Fig. 2). In terms of heterogeneity, it was statistically significant, and the amount of heterogeneity according to the I Square index was equal to 69.44%. According to the Eggers test and funnel plot diagram, the publication bias occurred in the pooled prevalence of depression in women with ovarian cancer (B = 9.36; SE = 3.11; P = 0.002; Fig. 3a). The meta-regression results showed that with the increasing age of the affected women, the prevalence of depression also increased, but it was not statistically significant (B = 0.05; SE = 0.01; P = 0.634; Fig. 3b).
The results of subgroup analysis based on the variables of the country, age, and NOS score were reported in Table 2. Combining six studies [13, 15, 20, 25, 26, 30] examining 1169 Australian women with ovarian cancer showed a pooled prevalence of depression of 12% (95% CI: 1% - 23%) in Australia. Two studies [16, 17] were performed on 192 women with ovarian cancer in the United States. After combining these results, the pooled prevalence of depression in American women with ovarian cancer was 36% (95% CI: 10% - 64%). Two studies [22, 23] with a sample size of 468 people were conducted in China. After combining these studies, the pooled prevalence of depression in Chinese women with ovarian cancer was 47 (95% CI: 42% - 53%). Based on the different scores of the NOS quality assessment checklist, the results of the subgroup analysis showed that the studies with NOS score of 6 were two studies with a sample size of 107 people. After combining these studies, the pooled prevalence of depression in women with ovarian cancer was 11% (95% CI: 4% - 18%). Also, studies with a high-quality evaluation score (equivalent to 8) were 11 studies with a sample size of 2184 people. After combining these articles, the pooled prevalence of depression in women with ovarian cancer was 35% (95% CI: 15% - 56%).
Subgroup analysis based on age showed that six studies [14, 15, 23, 24, 28, 29] with a sample size of 1280 people determined the prevalence of depression in women with ovarian cancer, aged equal to or less than 55 years. After combining these results, the pooled prevalence of depression was estimated as 30% (95% CI: 12% - 50%) in women with ovarian cancer, aged equal to or less than 55 years. Also, the prevalence of depression in women with ovarian cancer older than 55 years, after combining 11 studies [13, 16-18, 20-22, 25-27, 30] with a sample size of 1704 people, was estimated as 26% (95% CI: 11% - 41%).
Prevalence of anxiety in women with ovarian cancer
In this section, 16 studies [13-19, 22-30] determined the prevalence of anxiety in women with ovarian cancer, with the lowest prevalence in the survey of Camara C et al. [27] with a value of 5% (95% CI: 1% - 8%) and the highest prevalence in the study of Stafford L et al. [25] with a value of 77% (95% CI: 68% - 87%). In these studies, a total of 2442 women with ovarian cancer were studied, of whom 737 people suffered from an anxiety disorder. After combining the results of these studies, the pooled prevalence was 33% (95% CI: 21% - 44%; Fig. 4). However, the amount of heterogeneity was statistically significant, and according to the I Square index, it was equal to 78.55%. Egger test and the funnel plot diagram showed that publication bias occurred in combination with the results of studies to determine the pooled prevalence of anxiety in women with ovarian cancer (B: 6.48; SE: 2.81; P: 0.021; Fig. 5a). The meta-regression results showed that with the increasing age of the affected women, the prevalence of anxiety also decreased, but it was not statistically significant (B: -0.02; SE: 0.01; P: 0.834; Fig. 5b).
The results of subgroup analysis based on the variables of the country, age, and NOS score were reported in Table 2. Five studies [13, 15, 25, 26, 30] determined the prevalence of anxiety in Australian women with ovarian cancer and examined a total of 556 people. The pooled prevalence of anxiety in Australian women with ovarian cancer was 30% (95% CI: 5% - 55%). Two studies [16, 17] were performed on 192 women with ovarian cancer in the United States. After combining these results, the pooled prevalence of anxiety in American women with ovarian cancer was 13% (95% CI: 1% - 26%). Two studies [22, 23] with a sample size of 468 people were conducted in China. After combining these studies, the pooled prevalence of anxiety in Chinese women with ovarian cancer was 55% (95% CI: 49% - 61%). Two studies [18, 19] with a sample size of 148 women with ovarian cancer were performed in the UK. After combining these results, the pooled prevalence of anxiety in British women with ovarian cancer was 43% (95% CI: 19% - 47%). The studies with the NOS score of 6 were two studies with a sample size of 30 people. After combining these studies, the pooled prevalence of anxiety in women with ovarian cancer was equal to 28% (95% CI: 20% - 37%). Also, studies with a high-quality NOS score (equal to 8) were 9 studies with a sample size of 517 people. After combining these studies, the pooled prevalence of anxiety in women with ovarian cancer was 42% (95% CI: 20% - 63%).
Six studies [14, 15, 23, 24, 28, 29] with a sample size of 1280 people determined the prevalence of anxiety in women with ovarian cancer, aged equal to or less than 55 years. After combining these results, the prevalence of anxiety in women with ovarian cancer, aged equal to or less than 55 years, was 39% (95% CI: 21% - 57%). Also, the prevalence of anxiety in women with ovarian cancer older than 55 years, after combining ten studies [13, 16-19, 22, 25-27, 30] with a sample size of 1091 people, was equal to 29% (95% CI: 13% - 45%).