Diagnostic delay of myositis: an integrated systematic review

Out of 328 studies identified, 76 duplicates were removed and 206 were excluded at title and abstract screening. A further 19 studies were excluded at full-text screening. The remaining 27 studies published between 1992 and 2020 were included in the review as described in Fig. 1.

Fig. 1figure 1

Study selection flow chart

Description of included studies

The descriptive summary of the selected studies is presented in Table 2. A full data extraction table is shown in Additional file 1: Table 4. The 27 studies included in the review consisted of seventeen non-comparative studies (nine descriptive cross-sectional studies [9, 14, 16, 17, 22, 23, 29, 36, 38], seven case reports [27, 28, 31,32,33,34,35] and one case series [30]), six analytical cross-sectional studies [15, 18,19,20, 24, 37], one time series with comparison group [21] and three retrospective cohort studies [25, 26, 41]. Studies were from multiple countries, including nine from Europe [9, 15, 16, 25, 27, 31, 35, 37, 38], eight from the United States [20, 21, 24, 29, 30, 33, 34, 41], three from Australia [19, 22, 23], one from New Zealand [14], two from Asia [28, 32], and one from South America [26]. Two studies were multi-national [18, 36]. Studies reported diagnostic delay in all IIM subtypes; DM [28, 29, 33] (n = 3 studies), IBM [9, 14, 16, 17, 21,22,23,24, 27, 30, 34,35,36] (n = 13 studies), ASS [18, 26, 31, 32] (n = 4 studies), JDM [15, 20] (n = 2 studies), NM [41] (n = 1 study) and mixed or all types of IIM [19, 21, 25, 38] (n = 4 studies). In total, based on the IIM subtypes, the systematic review included a sample size of 1827 with the highest sample size for IBM (n = 1262 people) [9, 14, 16, 21,22,23,24, 27, 30, 34,35,36] and the lowest for NM [41] (n = 67 people).

Table 2 Descriptive summary table of selected studies (n = 27)Pooled diagnostic delay in IIM

Nineteen studies were included in the meta-analysis with a total sample size of 1518 people. Individual study sample size for the 19 studies range from six [14] to 478 [25]. The mean diagnostic delay in IIM ranged from 3.48 [25] to 96.0 months [16]. The pooled overall mean diagnostic delay was 27.91 months (95% CI 15.03–40.79, I2 = 99%) as outlined in Fig. 2. The funnel plot for overall mean diagnostic delay is shown in Additional file 1: Figure S2. Excluding studies with estimated SD showed similar results as summarised in Additional file 1: Figure S3.

Fig. 2figure 2

Forrest plot for mean diagnostic delay in all types of IIM

Subgroup analyses

Subgroup analyses between different types of IIM is summarised in Fig. 3 (n = 19). Subgroup analyses revealed significant differences in mean diagnostic delay between different IIM subtypes. Compared to other IIM subtypes JDM had the shortest mean delay (6.73 months, 95% CI = − 10.40–23.85) whereas IBM had the longest mean delay (61.95 months, 95% CI = 47.66–76.24).

Fig. 3figure 3

Forrest plot for all types of IIM

Subgroup analyses between IBM and non-IBM types are summarised in Fig. 4 (n = 19). There was a significant difference in mean diagnostic delay between IBM and non-IBM types. Compared to non-IBM (12.52 months, 95% CI = 3.89–21.15), IBM type had significantly longer mean diagnostic delay (61.32 months, 95% CI = 44.99–77.65).

Fig. 4figure 4

Forrest plot for mean diagnostic delay in IBM and non-IBM types of IIM

Subgroup analysis between studies for which MSA was tested versus studies for which MSA was not tested is presented in Additional file 1: Figure S4 (n = 19). The mean diagnostic delay did not differ between ‘MSA-tested’ and ‘MSA-not-tested’ studies.

Subgroup analysis between studies conducted in gatekeeper health systems versus non-gatekeeper health systems is summarised in Fig. 5 (n = 18). Countries with a gatekeeper health system had significantly longer diagnostic delay (34.37 months, 95% CI = 13.19–55.56) compared to countries with a non-gatekeeper health system (27.42 months, 95% CI = 5.60–49.24).

Fig. 5figure 5

Subgroup analysis between gatekeeper and non-gatekeeper health systems

Subgroup analysis between studies that used Peter and Bohan’s criteria and ENMC criteria is presented in Additional file 1: Figure S5 (n = 5). There was no significant difference in mean diagnostic delay in studies where Peter and Bohan’s criteria and those where the ENMC criteria was used.

Subgroup analysis comparing studies at multidisciplinary centres and specialist centres is presented in Additional file 1: Figure S6 (n = 16). The mean diagnostic delay did not differ between different centres.

Case studies

There were eight case studies, including seven case reports [27, 28, 31,32,33,34,35] and one IBM case series [30]. The seven case reports consisted of three IBM [27, 34, 35], two ASS [31, 32], and two DM case studies [28, 33]. These included an overall sample size of 27 people (17 males, 10 females, mean age = 56.41 years, (95% CI = − 73.68 to 186.60)). Mean diagnostic delay in case studies was 49.87 months (95% CI = 17.92–81.83) ranging from 0 months for ASS [32] and 120 months for DM [33].

Initial specialist, initial symptoms, most common symptoms, and symptoms that changed the diagnosis

Four of the included studies reported the initial specialist seen by patients [17, 30,31,32] as described in Additional file 1: Table S4. Due to the lack of available data, further analysis on the difference in diagnostic delay related to the initial specialist seen could not be conducted.

Eighteen of the included studies reported a broad range of initial symptoms as shown in Additional file 1: Table S5. Due to heterogenous data, investigation of associations between the initial symptoms and the diagnostic delay was not conducted. However, further analysis of initial symptoms with meta-aggregation by each subtype of IIM is presented in Additional file 1: Table S5. Based on the case reports, 50% (3/6) of the reported symptoms for ASS were lung symptoms, and for IBM 83.3% of initial symptoms were muscle associated symptoms (5/6 symptoms). In JDM, based on a single time series [18] and one cross-sectional study [20], the most common initial symptoms were muscle symptoms (proximal muscle weakness (93%), muscle pain (31–75%)), skin symptoms (Gottron’s papules (74%), heliotrope rash (67%), periungual/nailfold capillary changes (35–75.6%), malar erythema (48%)), and general symptoms including (fatigue (44–82%), weight loss (33–44%), and arthralgia (40–61%)).

Nine studies consisting of two retrospective cohort studies [21, 25], five cross sectional studies [15, 18,

留言 (0)

沒有登入
gif