3.4 Outcome measures
3.4.1 Pooled prevalence of PD in the acute and chronic phases after aSAH
Data from 14 literatures[8, 14, 19, 21, 22, 29, 34–36, 39, 42, 43, 45, 46] assessed the PD after aSAH in the acute phase (Fig. 2). The prevalence of PD after aSAH within 6 months was 0.50, which was estimated to vary from 0.32 to 0.67 (I2 = 98.0%, P value <0.001), with a total of 1148 individuals. The subtotal prevalence of PD was 0.59 (95% CI, 0.44–0.75, I2 = 95.9%, P value <0.001) within 3 months and 0.23 (95% CI, 0.13–0.33, I2 = 75.6%, P value = 0.006) in 3–6 months. As shown in Fig. 2, it can be seen that the 95% confidence interval of the prevalence within 3 months and 3–6 months has no overlap, so the prevalence of PD within 3 months is significantly higher than that within 3–6 months (P value<0.001).
Similarly, 20 literatures’ data[8, 10, 13, 19, 21, 23, 30, 31, 33–40, 42, 44, 45, 47] with a total of 1453 subjects evaluated PD after aSAH in the chronic phase (Fig. 3). The prevalence of PD after aSAH later than 6 months was 0.30, which estimated ranged from 0.21 to 0.39 (I2 = 94.7%, P value <0.001). The subtotal prevalence of PD was 0.29 (95% CI, 0.12–0.46, I2 = 97.5%, P value <0.001) during 6–12 months. When assessed after 12 months, the subtotal prevalence of PD was 0.31 (95% CI, 0.22–0.41, I2 = 86.5%, P value <0.001). We found the prevalence rates of PD increased over time after 6 months but were not statistically significant (P = 0.817), indicating that the increase in prevalence was not very obvious.
3.4.2 Global analysis for PD
By comprehensive analysis of the literature we included, we calculated the prevalence of each hormone deficiency in the acute and chronic phases.
In the acute phase, 15 studies[14, 19, 21, 22, 29, 32, 34–36, 39, 41, 43, 45, 46] with a total of 739 participants evaluated the prevalence of ACTH deficiency which was 0.15 (95% CI, 0.09–0.21, I2 = 90.0%, P value <0.001). The prevalence of GHD was 0.36 (95% CI, 0.21–0.51, I2 = 94.6%, P value <0.001) which evaluated by 12 studies[14, 19, 21, 29, 34–36, 39, 43, 45, 46] with a total of 632 participants. A total of 786 participants from 15 studies[14, 19, 21, 29, 32, 34–36, 39, 41–43, 45, 46] assessed the prevalence of TSH deficiency that was 0.17 (95% CI, 0.09–0.24, I2 = 94.7%, P value <0.001). Of 13 studies[14, 19, 21, 29, 34–36, 39, 42, 43, 45, 46] with a total of 705 participants evaluated the prevalence of Gn deficiency that was 0.33 (95% CI, 0.21–0.44, I2 = 93.4%, P value <0.001). About hyperprolactinemia, the prevalence was 0.12 (95% CI, 0.07–0.16, I2 = 69.3%, P value = 0.001) calculated by 12 studies[14, 19, 21, 29, 34, 35, 39, 42, 43, 45, 46] with a total of 621 participants. As shown in Table 3, it can be seen that the 95% confidence intervals of the prevalence of GHD or Gn deficiency and the remaining ACTH deficiency or hyperprolactinemia have no overlap respectively, so the prevalences of GHD or Gn deficiency were significantly higher than that of ACTH deficiency and hyperprolactinemia.
Table 3
Global Analysis and Subgroup Analysis of Pituitary Dysfunction in the acute phases CI: confidence interval, NA: not applicable.
Variable | No. of Articles | No. of Cases | No. of Participants | Prevalence (95% CI) | Heterogeneity | Sensitivity analysis | Subgroup difference |
Q test | I2, % | | | |
Global Analysis for classification of PD |
ACTH deficiency | 15 | 121 | 739 | 0.15 (0.09, 0.21) | P<0.001 | 90.03% | | | NA |
GH deficiency | 12 | 215 | 632 | 0.36 (0.21, 0.51) | P<0.001 | 94.57% | | | NA |
TSH deficiency | 15 | 124 | 786 | 0.17 (0.09, 0.24) | P<0.001 | 94.66% | | | NA |
Gn deficiency | 13 | 238 | 705 | 0.33 (0.21, 0.44) | P<0.001 | 93.37% | | | NA |
Hyperprolactinemia | 12 | 64 | 621 | 0.12 (0.07, 0.16) | P = 0.001 | 69.28% | | | NA |
Subgroup analysis of PD |
WHO region | | | | | | | | | P<0.001 |
ARFO | none | none | none | none | none | none | | |
PAHO | 2 | 79 | 148 | 0.54 (0.46, 0.62) | NA | NA | | |
SEARO | 2 | 115 | 173 | 0.81 (0.77, 0.86) | NA | NA | | |
EURO | 11 | 301 | 827 | 0.47 (0.28, 0.65) | P<0.001 | 97.07% | | |
EMRO | none | none | none | none | none | none | | |
WPRO | none | none | none | none | none | none | | |
Type | | | | | | | | | P = 0.674 |
Single | 11 | 175 | 606 | 0.28 (0.20, 0.35) | P<0.001 | 78.55% | | |
Multiple | 11 | 175 | 606 | 0.25 (0.12, 0.37) | P<0.001 | 95.04% | | |
In the chronic phase, the prevalence of ACTH deficiency was 0.21 (95% CI, 0.12–0.29, I2 = 91.21%, P value <0.001) which evaluated by 19 studies[10, 13, 19, 21, 23, 31–33, 35–40, 45, 47] with a total of 880 participants. Of 22 studies[10, 13, 19, 21, 23, 30–40, 44, 45, 47] with a total of 1018 participants evaluated the prevalence of GHD that was 0.18 (95% CI, 0.14–0.22, I2 = 63.40%, P value <0.001). About TSH deficiency, the prevalence was 0.05 (95% CI, 0.02%-0.07%, I2 = 41.94%, P value = 0.070) calculated by 22 studies[10, 13, 19, 21, 23, 31–40, 42, 44, 45, 47] with a total of 1062 participants. Of 22 studies[10, 13, 19, 21, 23, 31–40, 42, 44, 45, 47] with a total of 1062 participants evaluated the prevalence of Gn deficiency that was 0.14 (95% CI, 0.09–0.19, I2 = 82.57%, P value <0.001). 19 studies[13, 19, 21, 23, 31–35, 37–40, 42, 44, 45, 47] assessed hyperprolactinemia after 6 months. This represents data from 918 participants, and the prevalence was 0.03 (95% CI, 0.01–0.04, I2 = 0.00%, P value = 0.481). Among these hormone deficiencies, the 95% confidence intervals of the prevalence of ACTH deficiency or GHD or Gn deficiency and the remaining TSH deficiency or hyperprolactinemia have no overlap apart, which can be considered that the prevalences of the three hormone deficiencies were higher than that of TSH deficiency or hyperprolactinemia (Table 4). The amount of literature on diabetes insipidus was too small, so no further analysis was performed after statistics.
Table 4
Global Analysis and Subgroup Analysis of Pituitary Dysfunction in the chronic phases
Variable | No. of Articles | No. of Cases | No. of Participants | Prevalence (95% CI) | Heterogeneity | Sensitivity analysis | Subgroup difference |
Q test | I2, % | | | |
Global Analysis for classification of PD |
ACTH deficiency | 19 | 124 | 880 | 0.21 (0.12, 0.29) | P<0.001 | 91.21% | | | NA |
GH deficiency | 22 | 180 | 1018 | 0.18 (0.14, 0.22) | P<0.001 | 63.40% | | | NA |
TSH deficiency | 22 | 33 | 1062 | 0.05 (0.02, 0.07) | P = 0.070 | 41.94% | | | NA |
Gn deficiency | 22 | 106 | 1062 | 0.14 (0.09, 0.19) | P<0.001 | 82.57% | | | NA |
Hyperprolactinemia | 19 | 22 | 918 | 0.03 (0.01, 0.04) | P = 0.481 | 0.00% | | | NA |
Subgroup analysis of PD |
WHO region | | | | | | | | | P<0.001 |
ARFO | none | none | none | none | none | none | | |
PAHO | 1 | 17 | 68 | 0.25 (0.15, 0.37) | NA | NA | | |
SEARO | 2 | 21 | 133 | 0.05 (0.02, 0.09) | NA | NA | | |
EURO | 18 | 396 | 1219 | 0.33 (0.24, 0.43) | P<0.001 | 93.12% | | |
EMRO | none | none | none | none | none | none | | |
WPRO | 1 | 5 | 33 | 0.15 (0.05, 0.32) | NA | NA | | |
Type | | | | | | | | | P<0.001 |
Single | 18 | 223 | 892 | 0.24 (0.16, 0.31) | P<0.001 | 90.73% | | |
Multiple | 18 | 55 | 892 | 0.07 (0.05, 0.10) | P = 0.062 | 43.19% | | |
CI: confidence interval, NA: not applicable. |
3.4.3 Subgroup analysis for PD
For the WHO Regional Office of PD after aSAH, the studies we included were located at EURO, SEARO, WPRO, and PAHO analyzing the prevalence of PD, respectively. Studies from the remaining two regions were not available.
Of the 15 studies, 11 studies[8, 14, 19, 21, 22, 29, 34–36, 39] with a total of 827 participants calculated the prevalence of PD to be 0.47 (95% CI, 0.28–0.65, I2 = 97.07%, P value <0.001) in the EURO. For the rest of 4 studies, two studies[45, 46] from PAHO with a total of 148 participants had a calculated prevalence of 0.54 (95% CI, 0.46–0.62), and the other two studies[42, 43] from SEARO had a calculated incidence of 0.81 (95% CI, 0.77–0.86) with a total of 173 participants. We found statistical significance among three regions (Table 3, P values<0.001) about the prevalence of PD in the acute phase (Table 3).
Of the 22 studies, 18 studies[8, 10, 13, 19, 21, 23, 30, 31, 33–40] with a total of 1219 participants calculated the prevalence of PD to be 0.33 (95% CI, 0.24–0.43, I2 = 93.12%, P value <0.0001) in the EURO. For the rest of 4 studies, two studies[42, 44] from SEARO with a total of 133 participants had a calculated prevalence of 0.05 (95% CI, 0.02–0.09), and one study[45] from PAHO had a calculated incidence of 0.25 (95% CI, 0.15–0.37) with a total of 68 participants, and the other one study[47] from WPRO had a calculated incidence of 0.15 (95% CI, 0.05–0.32) with a total of 33 participants. We found statistical significance among four regions (Table 4, P values<0.001) about the prevalence of PD in the chronic phase (Table 4).
In the acute phase, the random effects pooled meta-analysis performed on 11 studies with a total of 1212 participants showed an overall prevalence of single pituitary hormone dysfunction[14, 19, 29, 34–36, 39, 43, 45, 46] of 0.28 (95% CI, 0.20–0.35, I2 = 78.55%, P value<0.0001) and an overall prevalence of multiple pituitary hormone dysfunctions[14, 19, 29, 34–36, 39, 43, 45, 46] of 0.25 (95% CI, 0.12–0.37, I2 = 95.04%, P value<0.0001), respectively. As shown in Table 3, it can be seen that the prevalence of single pituitary hormone dysfunction was higher than that of multiple pituitary hormone dysfunction, but not statistically significant.
In the chronic phase, the random effects pooled meta-analysis performed on 18 studies with a total of 1784 participants showed an overall prevalence of single pituitary hormone dysfunction[10, 13, 23, 30, 31, 33–40, 42, 45, 47] of 0.24 (95% CI, 0.16–0.31, I2 = 90.73%, P value<0.0001) and an overall prevalence of multiple pituitary hormone dysfunction[10, 13, 23, 30, 31, 33–40, 42, 45, 47] of 0.07 (95% CI, 0.05–0.10, I2 = 43.19%, P value = 0.062), respectively. As shown in Table 4, it can be seen that the 95% confidence interval of the prevalence of single pituitary hormone dysfunction and multiple pituitary hormone dysfunctions has no overlap, so the prevalence of single pituitary hormone dysfunction is significantly higher than that of multiple pituitary hormone dysfunctions.