Diagnostic delay in axial spondyloarthritis: a systematic review and meta-analysis

Abstract

Background

Delay to diagnosis in axial SpA (axSpA) is longer than in many other rheumatic diseases. Prolonged delay is associate with poorer outcomes, including functional impairment and quality of life. Our aims were to describe global variation in delay to diagnosis, factors associated with delay, and delay compared with PsA.

Methods

We searched MEDLINE, PubMed, Embase and Web of Science using a predefined protocol. Diagnostic delay was defined as years between the age at symptom onset and at diagnosis. We pooled the mean delay using random effects inverse variance meta-analysis. We examined variations in pooled estimates using prespecified subgroup analyses and sources of heterogeneity using meta-regression.

Results

A total of 64 studies reported the mean diagnostic delay in axSpA patients. The pooled mean delay was 6.7 years (95% CI 6.2, 7.2) with high levels of heterogeneity. Delay to diagnosis did not improve over time when stratifying results by year of publication. Studies from high-income countries (defined by the World Bank) reported longer delays than those from middle-income countries. Factors consistently reported to be associated with longer delays were lower education levels, younger age at symptom onset and absence of extra-articular manifestations (EAMs). The pooled estimate for diagnostic delay from 8 PsA studies was significantly shorter, at 2.6 years (95% CI 1.6, 3.6).

Conclusion

For axSpA patients, delay to diagnosis remains unacceptably prolonged in many parts of the world. Patient factors (e.g. education) and disease presentation (onset age and EAMs) should inform campaigns to improve delay.

axial spondyloarthritis, ankylosing spondylitis, delay to diagnosis, psoriatic arthritis, meta-analysis

Rheumatology key messages

• The mean delay to diagnosis in axial SpA is 6.7 years worldwide.

• This is significantly longer than the mean delay of 2.6 years reported in PsA studies.

• Lower education, absence of extra-articular manifestations and younger age of onset are associated with delays in diagnosis of axSpA.

Introduction

Axial SpA (axSpA) is a chronic inflammatory disease characterized by significant inflammatory pain, stiffness and functional impairment [1]. Symptoms typically begin in early adulthood, which is a critical time for education, career, social networks and development of personal identity in general. Consequently, axSpA can significantly impact on mental health, quality of life and work productivity over the life course, at costs to the individual and the economy [2, 3].

The disease impact is often compounded by a prolonged diagnostic delay, that is, the time from onset of symptoms to getting a diagnosis. This may be explained by the lack of awareness of axSpA as a cause of back pain among patients and primary care providers, and/or suboptimal referral to appropriate specialists or for investigations. The duration of delay is reported to range from 8 to 10 years—longer than for many other rheumatic diseases—although estimates can vary considerably between studies. Some studies have also found no improvement in diagnostic delay over recent decades [4], despite improved understanding of the disease and access to imaging.

There is abundant evidence that diagnostic delay is associated with worse functional impairment, greater radiographic progression, poorer quality of life and reduced response to treatment [5, 6]. Those with longer delay to diagnosis also report greater work disability, unemployment and healthcare costs [5]. Although the impact of delay is well described, potential causes of delay (i.e. how delay can be improved) are not. Examining how delay durations vary across parts of the world and factors associated with delay will help inform targets for improvement.

The aims of this systematic review were to describe global variation in, and factors associated with, diagnostic delay in axSpA. We also compared delay duration in axSpA with other SpAs (e.g. psoriatic arthritis) to highlight the need for improvement.

Methods

We performed a systematic review and reported it in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [7]. The protocol for this review was registered in advance (PROSPERO: CRD42020161887). We searched MEDLINE, PubMed, Embase and Web of Science for relevant literature in September 2019 using the following search terms: (ankylosing OR spondyloarthritis OR psoriatic) AND ((delay AND diagnosis) OR (symptom AND (onset OR duration))). This search strategy was designed to include studies that describe diagnostic delay, even if delay was not their research objective. In response to peer review, bibliographies of all eligible studies and a prior systematic review [8] were also manually searched to identify additional titles.

Studies were included if they reported the mean delay to diagnosis (i.e. the mean difference between the age at symptom onset and at diagnosis) or if they reported both mean age at onset and at diagnosis. Studies that defined eligibility using these variables were not eligible since estimates would not be representative. We excluded studies using the same (or very similar) cohort as studies already included. We also excluded studies reporting medians only. Letters and published conference abstracts were considered, as some prevalence studies may have sufficiently detailed methodology and results but not published as full articles. Non-systematic reviews, comments and editorials were excluded.

Two independent reviewers (B.P., N.H.) screened titles and abstracts, assessed full texts for eligibility and extracted data from qualifying studies. Any discrepancy at each stage was resolved through discussion moderated by a third reviewer. Information from included studies was extracted into predefined tabulated summaries (Supplementary Table S1). Studies were assessed for risk of bias using adapted versions of the Newcastle–Ottawa Scale (Supplementary Table S2).

Analysis

We pooled mean diagnostic delay using inverse variance weighted random effects models (DerSimonian–Laird method). This was performed for studies of axSpA (including AS), then separately for PsA and SpA (which includes axSpA, PsA and other members of the SpA family). Where the mean delay to diagnosis was not reported, it was imputed as the difference in mean age at symptom onset and the mean age at diagnosis. Where the standard deviation of diagnostic delay was missing, we imputed it using methods recommended by Cochrane (in essence, based on standard deviations of age at onset, age at diagnosis and their correlation in all studies [9]) or the standard deviation of a similar sized study reporting the most similar mean delay duration. We performed sensitivity analyses without imputed values. Further sensitivity analyses requested at peer review were performed by restricting to studies using classification criteria only and excluding studies with potentially non-representative sampling (e.g. entirely male populations or sample sizes of <30 that may give unstable population estimates). Heterogeneity of meta-analysis estimates was presented using the I² statistic. Funnel plots were used to assess the risk of publication bias.

We used random effects meta-regression to examine whether heterogeneity in axSpA diagnostic delay could be explained by study characteristics, including year of publication (pre-2010, 2010–2015, post-2015), geography (regions defined by the World Health Organization (WHO) [10]), economic status of the country (World Bank economic class [11]), sample sources (e.g. single centre, multicentre), age at symptom onset (tertiles) and proportion of males (tertiles). Meta-regression was not performed for PsA and SpA due to the limited number of studies. Analyses were performed using R version 3.6.2 (R Foundation for Statistical Computing, Vienna, Austria) and the ‘meta’ and ‘metafor’ packages.

Results

A total of 3286 publications were found from the literature search. After excluding duplicates, irrelevant or ineligible studies and studies using the same cohorts (or subsets thereof), 66 studies remained. A further 12 studies were found through manual bibliography searches. One study was excluded, as it only included patients with ≤2 years of symptoms. The selection flowchart is shown in Supplementary Fig. S1. The 78 included studies are summarized in Supplementary Table S1; 64 studies reported delays among axSpA patients, 8 in PsA and 8 in SpA. Feld et al. [7] and Sørensen et al. [12] reported delays in both axSpA and PsA. Bias scores were mostly 3–4 out of 6 stars (Supplementary Table S2 and Fig. S2), indicating moderate bias.

Diagnostic delay in axSpA

The sample size for axSpA studies ranged from 5 to 2887 patients. A total of 47 studies were of AS (including 32 using modified New York criteria) and 17 of axSpA [including 11 using Assessment of SpondyloArthritis international Society (ASAS) criteria)]. Delay ranged from 2.8 years in a small Albanian study (of 54 cases over 6 years) to 11.1 years in a single UK centre [13, 14]. The mean delay to diagnosis was 6.7 years overall (95% CI 6.2, 7.2; I² = 99%) (Fig. 1). The accompanying funnel plot is shown in Supplementary Fig. S3.

Fig. 1

Pooled estimate of diagnostic delay in axSpA (including AS)

Results ordered according to geography and year of publication. *Diagnostic delay calculated from summary data for age at onset and diagnosis. **Standard deviation imputed for meta-analysis.

Results of stratified meta-analysis are shown in Table 1. A total of 39 axSpA studies were from countries in the European region, 9 in the West Pacific, 8 in the Eastern Mediterranean, 5 in the Americas and 3 in Southeast Asia. Across these WHO regions, the mean delay and heterogeneity were not significantly different. When these studies were stratified according to World Bank economic class, the high-income group had longer mean delays than the upper- and lower-middle-income countries. When mean delays were pooled according to country (with three or more studies), the average diagnostic delay was significantly shorter in Turkey and China than in the UK. The mean delay duration did not differ according to the year of publication. Studies with an older mean age of symptom onset showed trends for shorter delay durations.

Table 1

Meta-analysis of delay duration stratified by study characteristics

	n	Mean delay	95% CI	I2, %
WHO regions
European	39	7.09	6.46, 7.72	97.9
West Pacific	9	5.92	4.34, 7.51	96.8
Eastern Mediterranean	8	6.27	4.92, 7.61	97.7
Americas	5	6.36	3.70, 9.03	99.8
Southeast Asia	3	6.38	0, 14.2	96.7
World Bank economic class
High	39	7.56	7.04, 8.09	97.5
Upper middle	20	5.37	4.54, 6.20	96.4
Lower middle	5	5.59	3.21, 7.96	95.7
Countries with three or more studies
UK	9	8.65	7.35, 10.0	94.4
Turkey	10	5.88	4.80, 6.96	90.3
Italy	3	7.68	2.67, 12.69	99.6
Iran	4	6.44	3.31, 9.58	98.2
China	4	4.32	2.63, 6.00	79.1
Recruiting methods
Single centre	37	6.45	5.74, 7.16	99.0
Multicentre	27	7.14	6.45, 7.84	99.2
Year of publication
<2010	12	6.75	5.66, 7.83	99.6
2010–15	30	6.83	6.05, 7.61	97.1
>2015	22	6.61	5.70, 7.53	99.3
Age at symptom onset (tertiles), years
22.7–24.0	12	7.01	5.69, 8.33	97.7
24.2–27.1	13	7.11	6.04, 8.19	98.5
27.1–35	13	6.25	5.03, 8.23	99.0
Proportion of males (tertiles), %
39–68	20	6.95	6.08, 7.82	98.5
69–80	19	7.30	6.31, 8.29	98.2
81–100	19	5.65	4.91, 6.39	94.5

The WHO regions, economic class, recruiting centre, year of publication, age at symptom onset and male proportion were entered into a multivariable meta-regression model (Supplementary Table S3). Compared with high-income countries, those in the upper- (by 2.5 years; P < 0.01) and lower-middle-income (3.7 years; P = 0.03) category had a shorter mean delay. Compared with studies from Europe, those from the Americas (by 2.1 years; P = 0.07) and West Pacific region (2.9 years; P = 0.01) had a shorter mean delay.

Sensitivity analysis restricting to 43 studies using classification criteria showed a similar mean delay duration of 6.5 years (Supplementary Fig. S4). The diagnostic delay was 6.3 years (95% CI 5.6, 7.0) for modified New York AS criteria and 7.1 years (95% CI 5.5, 8.7) for ASAS axSpA criteria. Excluding three studies with potentially non-representative sampling (n < 30 or all-male populations) did not change the results (data not shown). Sensitivity analyses excluding studies with an imputed mean and/or standard deviation of delay produced similar results (data not shown).

Factors associated with delay to diagnosis

Most studies performed unadjusted comparisons (Table 2). Delay was reportedly longer in males in studies by Bandinelli et al. [15] (10 vs 6.3 years; P = 0.002) and Sykes et al. [4] (9.4 vs 8.3; P = 0.097), but longer in females in studies by Fallahi and Jamshidi [16] (8.7 vs 7.7; P = 0.68), Dincer et al. [17] (14 vs 5.3; P = 0.06), Hajialilo et al. [18] (8.0 vs 5.9; P = 0.14), Jones et al. [19] (8.5 vs 5.6) and Redeker et al. [20] (by 1.9 years; P < 0.05), although mostly not statistically significant. Similarly, two studies reported a longer delay in those with peripheral arthritis [16, 18], while five reported longer delays in those without [4, 6, 15, 21, 22]. There was also inconsistency in whether studies found HLA-B27 status to be associated with diagnostic delay: 4 studies reported significantly longer delays in HLA-B27-negative patients [16, 17, 20, 23], while five other studies did not [15, 21, 22, 24, 25].

Table 2

Factors associated with longer delay to diagnosis in axSpA (results reported as mean duration in years)

Aggarwal and Malaviya 2009 [24]	Absence EAMs vs presence (8.7 vs 5.9; P = 0.03) Onset <16 vs ⩾16 years (9.1 vs 6.1; P = 0.03)
Bandinelli et al. 2016 [15]	Males vs females (10 vs 6.3; P = 0.002) Manual vs non-manual workers (11 vs 8.3; P = 0.047) Axial presentations compared with arthritis or enthesitis (9.0 vs 8.5 vs 4.3; P = 0.002) Lower education (<high school vs high school vs university: 10 vs 8.6 vs 7.3, P = 0.076)
Dincer et al. 2008 [17]	HLA-B27 negative vs positive (9.2 vs 5.3; P = 0.037) Family history vs none (10 vs 4.6; P = 0.003) Onset ≤16 vs >16 years (8.9 vs 5.5; P = 0.027) Lower education (<9 years vs 9–11 vs 12–13 vs 14–15: 12 vs 6.3 vs 5.0 vs 4.6; P = 0.018) Females vs males (14 vs 5.3; P = 0.061)
Fallahi and Jamshidi 2016 [16]	Enthesitis vs no enthesitis (8.8 vs 6.0; P = 0.007) HLA-B27 negative vs positive (10 vs 7.1; p = 0.013) Lower education (correlation r = 0.24, P = 0.002) Presence of peripheral arthritis vs absence (8.9 vs 6.8; P = 0.086)
Feldtkeller et al. 2003 [23]	HLA-B27 negative vs positive (11 vs 8.5, P < 0.01)
Gerdan et al. 2012 [26]	With vs without prior diagnosis of lumbar disc herniation (9.1 vs 6.2; P = 0.002) First contact being rheumatology vs non-rheumatology (8.1 vs 2.9; P < 0.001) Younger age at onset (β = −0.18, P = 0.003) Lower education (β = −0.252, P = 0.018)
Hajialilo et al. 2014 [18]	Presence of peripheral arthritis vs absence (11 vs 5.1; P < 0.001) Absence of uveitis vs presence (6.4 vs 2.4; P =  0.02) Presence of heal pain vs absence (13 vs 5.9; P = 0.004) Females vs males (8.0 vs 5.9; P = 0.14)
Jones et al. 2014 [19]	Females vs males (8.5 vs 5.6)
Masson Behar et al. 2017 [21]	Univariable regression showed longer delay with Older age at diagnosis (β = 0.15, P < 0.001) Lower education (β = −1.7, P = 0.03) Later calendar year of diagnosis (β = 0.1, P = 0.005) Multivariable regression showed longer delay with Older age at diagnosis (β = 0.1, P < 0.001) Entheseal pain vs none (β = 1.5, P = 0.015) Absence of peripheral arthritis/dactylitis vs presence (β = −1.7, P = 0.005)
Nakashima et al. 2016 [22]	Absence of articular involvement vs presence (8.9 vs 5.2, P = 0.03) Disease onset pre-2000 vs post (7.5 vs 3.5; P = 0.02)
Reed et al. 2008 [27]	Delay longer with later calendar year and younger age at onset (P < 0.05)
Seo et al. 2015 [6]	Long delay vs short delay (≤8 years) category associated with Absence of peripheral symptoms (OR 2.2, P = 0.06) Prior diagnosis of mechanical back pain (OR 2.8, P = 0.02) In univariate analysis, mechanical back pain remained significant in the multivariable model
Sykes et al. 2015 [4]	Absence of peripheral arthritis vs presence (9.4 vs 7.6; P = 0.045) Absence of IBD vs presence (9.2 vs 6.5; P = 0.012) Presence of uveitis vs absence (10 vs 8.4; P = 0.033) Females vs males (9.4 vs 8.3; P = 0.097)
Redeker et al. 2018 [20]	Multivariable regression showed longer delay in Females vs males (β = 1.9; 95% CI 1.1, 2.7) Younger age of symptom onset per 10 years [−1.9 (95% CI −2.3, −1.5)] HLA-B27 negative vs positive [−3.6 (95% CI −5.1, −2.1)] Psoriasis vs no psoriasis [1.4 (95% CI 0.1, 2.7)]
Resende et al. 2018 [25]	Presence of EAMs vs absence (8.7 vs 5.0, P < 0.001) Younger age at onset (r = −0.28, P < 0.001)

EAMs: extra-articular manifestations (anterior uveitis, psoriasis, inflammatory bowel disease).

There was better consensus among the studies that longer delay was associated with the absence of EAMs [4, 18, 24], lower educational attainment [16, 17, 21, 26] and younger age of onset [20, 21, 25, 26].

PsA and SpA

The sample size for 8 PsA studies ranged from 69 to 1970 patients. The diagnostic delay ranged from 1.0 years in the Dutch South-West Psoriatic Arthritis cohort to 4.6 in a Swedish population-based cohort [27, 28]. The mean delay to PsA diagnosis was 2.6 years (95% CI 1.6, 3.6; I² = 99%) (Fig. 2).

Fig. 2

Pooled estimate of diagnostic delay in PsA and SpA

*Diagnostic delay calculated from summary data for age at onset and diagnosis. **Standard deviation imputed for meta-analysis.

Eight SpA studies ranged from 16 to 708 participants in size and 1.6–7.6 years in diagnostic delay. The mean delay to SpA diagnosis was 4.9 years (95% CI 3.3, 6.6; I² = 96%) (Fig. 2). A funnel plot is shown in Supplementary Fig. S3.

Sensitivity analysis restricting to 3 PsA and 6 SpA studies using classification criteria showed similar results (Supplementary Fig. S5).

Discussion

The mean delay to diagnosis was 6.7 years across 64 axSpA studies worldwide. Factors associated with the delay to diagnosis varied and were often contradictory across studies; the most consistently reported factors were lower education, absence of extra-articular manifestations and younger age of onset. Diagnostic delay in axSpA was significantly longer than for PsA (2.6 years) and SpA (4.9 years).

The mean duration of delay varied significantly within (e.g. from 5.7 to 11 years in the UK) and between countries. This may reflect multiple factors that could not be assessed in this review, such as local healthcare infrastructure and awareness of the disease. Our finding that delay was longer in high-income countries was unexpected. It may be that research centres in these countries received referrals for the most diagnostically challenging cases or served comparably deprived areas. Conversely, it may be that only centres with good referral infrastructure are publishing research in middle-income countries.

Our meta-analysis showed no meaningful change in diagnostic delay over (publication) time. This is consistent with results from the UK [4, 29], France [21] and Germany [20]. In stark contrast, delay to diagnosis improved dramatically in Japan (pre- vs post-2000: 7.5 vs 3.6 years [22]), Italy (1990s vs 2000s: 7.4 vs 2.1 years [30]), Egypt (pre- vs post-2010: 11 vs 4.6 years [31]) and Australia [32]. We could not examine the cause of this variation in detail, but diagnostic approaches likely varied from country to country. For example, the extent to which HLA-B27 and gender were associated with delay differed between countries, suggesting that these factors may have differential importance in their respective diagnostic process.

Inflammatory back pain in axSpA typically has an insidious onset, with subtle signs on clinical examination. There is also a plethora of highly prevalent differential diagnoses that may be incorrectly used to explain symptoms; for example, lumbar disc disease can coexist with axSpA and prolong delay to diagnosis [6, 26]. Peripheral joint involvement is relatively more acute in presentation, with clearer signs such as swelling and erythema. This may explain the much shorter diagnostic delay in PsA than in axSpA. Among axSpA studies, the presence of peripheral joint involvement was associated with a shorter delay to diagnosis in Italian [15], UK [4], French [21] and Japanese [22] studies, while these patients had longer delays in Iran [16, 18]. It may be the case that these Iranian patients were given other diagnoses prior to the correct axSpA diagnosis.

To reduce delay to diagnosis, one potential target would be to improve general public awareness of axSpA as a cause of back pain, since general education was inversely associated with delay. Younger age of onset was also consistently associated with prolonged delay (although this may be an artefact of ‘delay’ being derived from and dependent on age at onset). AxSpA as a cause of chronic back pain in young people should receive greater emphasis among non-rheumatologists. Nevertheless, there will be cases that remain more diagnostically challenging, such as patients with few SpA features.

A key strength of this review is the large and globally representative number of studies. The unique search strategy allowed us to include studies that described diagnostic delay, even if delay was not their research objective. Such studies are less likely to be subject to subconscious bias from a prior delay hypothesis, thus their inclusion is a strength rather than a weakness. However, there were limitations. Diagnostic delay is known to be right-skewed in distribution, meaning that the mean is inflated above the median by a high proportion of people with disproportionately long delays. In other words, the mean may be sensitive to these outliers (e.g. atypical clinical features or individuals with poor access to healthcare) and remain unchanged, even if diagnostic delay generally improved for many patients. We chose mean because it permits meta-analysis, but also because median would take emphasis away from those with unusually long delays—precisely the individuals needing improvement to diagnosis. Some meta-analysis estimates for delay had negative lower bounds in the confidence interval. This is an artefact of the random effects methodology; in each case there is one study with a much shorter delay than others in the category, resulting in wide intervals required to cover the pooled estimate for this subgroup. This artefact disappears in fixed effects models, which were not used in this study due to high heterogeneity between the studies. We did not review the impact of delay to diagnosis, as this was recently reviewed [5]. Meta-regression examines relationships between summary data and should not be interpreted as traditional hypothesis testing of individual patient data. For example, the proportion of males was not associated with diagnostic delay, but this does not rule out a difference in delay between the sexes. Although most studies in our review did not report a statistically significant difference, a prior meta-analysis of SpA (excluding PsA) did [8]. Delay over time should also be interpreted with caution. The year of publication was the only available proxy for calendar time, since the recruitment period can be over many years and was often not reported. The intervals between recruitment and publication were generally homogeneous in studies that did report these data.

Conclusion

The delay from symptom onset to diagnosis in axSpA is 6.7 years on average, which is significantly longer than 2.6 years for PsA. Although delay has improved over time in some parts of the world, many countries such as the UK need additional efforts to improve delay to diagnosis. Lower educational attainment, absence of EAMs and younger age of onset were associated with longer delays, therefore improved education for physicians and patients with back pain may help reduce diagnostic delay.

Acknowledgements

We thank Karishma Nursiah for contributing to the design stage of the review and the UK National Axial Spondyloarthritis Society for their efforts to improve delay to diagnosis. S.Z. wrote the manuscript with significant contribution from all co-authors. B.P., N.L.H. and A.E.A. performed the literature search, data extraction and quality assessment. S.Z. and D.M.H. performed all statistical analysis and conceived of the project. All authors read and approved the final manuscript.

Funding: None of the authors received any funding related to the writing of this article.

Disclosure statement The authors have declared no conflicts of interest.

Data availability statement

All data relevant to the study are included in the article or uploaded as online supplementary information.

Supplementary data

Supplementary data are available at Rheumatology online.