Diagnostic Delay in Romanian Patients with Inflammatory Bowel Disease: Risk Factors and Impact on the Disease Course and Need for Surgery

Abstract

Background:

The epidemiology of inflammatory bowel disease [IBD] in Eastern Europe is poorly understood, particularly with regard to diagnostic delay. Here we investigated the factors leading to delayed diagnosis and the effect of the delay on several disease progression and outcome measures.

Methods:

A total of 1196 IBD cases [682 ulcerative colitis [UC], 478 Crohn’s disease [CD], 36 indeterminate colitis] from the Romanian national registry IBDPROSPECT were reviewed. Standard clinical and demographic factors were evaluated as predictors of a long diagnostic delay in both CD and UC. Diagnostic delay was subsequently evaluated as a potential risk factor for bowel stenoses, bowel fistulas, perianal fistulas, perianal surgery, and intestinal surgery in CD patients.

Results:

The median diagnostic delay was significantly longer in CD [5 months] than in UC [1 month] patients [p < 0.001]. Compared with 5 months for UC patients, 75% of CD patients were diagnosed within 18 months of symptom onset. In CD patients, extra-ileal location was a protective factor (odds ratio [OR], 0.5; p = 0.03), whereas being an active smoker [OR, 2.09; p = 0.01] and symptom onset during summer [OR, 3.35; p < 0.001] were independent risk factors for a long diagnostic delay [> 18 months]. In UC patients, an age > 40 years was a protective factor [OR, 0.68; p = 0.04] for a long delay. Regarding outcomes, a long diagnostic delay in CD patients positively correlated with bowel stenoses [OR, 3.38; p < 0.01] and any IBD-related surgery [OR, 1.95; p = 0.03] and had a positive trend for intestinal fistulas [OR, 2.64; p = 0.08] and perianal fistulas [OR, 2.9; p = 0.07]. Disease duration since diagnosis positively correlated with bowel stenoses [OR, 1.04; p = 0.04], any IBD-related surgery [OR, 1.04; p = 0.02], and intestinal surgery [OR, 1.07; p < 0.01].

Conclusions:

A long diagnostic delay in IBD correlates with an increased frequency of bowel stenoses and need for IBD-related surgery.

1. Introduction

Epidemiological data on inflammatory bowel disease [IBD] in Eastern European countries, Romania in particular, are scarce and have typically been obtained from retrospective, hospital-based surveys. However, in the past two decades, the quality of epidemiological studies in this area has improved considerably, evolving from retrospective reports to prospective cohort studies.¹ Despite the lack of national registries or population-wide epidemiological studies in developing countries, it seems that this pathology is not uncommon and its incidence is tending to increase in these parts of the world.²

The diagnosis of IBD, especially Crohn’s disease [CD], often presents a challenge for clinicians. The presenting symptoms [frequently of low grade] are sometimes mistaken for functional digestive pathologies [i.e. irritable bowel disease].³ A lower availability of diagnostic methods, particularly in rural areas, and a low clinical suspicion index, especially in developing countries, may both lead to a prolonged delay from the first occurrence of tell-tale symptoms to an established diagnosis.^4,5,6,7 Delayed diagnosis is often associated with lower patient quality of life, severe disease progression, poor responses to drug therapy, more complications, and a more frequent need for surgery.^8,9,10

The effects of demographic and environmental factors on the clinical presentation of IBD have been widely studied.^11,12 One of the goals of the present study was to identify the factors responsible for the delayed diagnosis of IBD. Another objective was to assess the impact a delay might have on the rate of complications and the need for IBD-related surgery.

2. Materials and Methods

2.1. Patients

In Romania, the diagnosis and therapeutic management of patients with IBD are mostly performed in referral centres. We have therefore implemented a web-accessible database [IBDPROSPECT] to facilitate the collection of IBD epidemiological data from all regions of Romania. The IBDPROSPECT project was initiated in 2006 by the partnership of two reference medical centres through a grant financed by the National Authority for Scientific Research and the Romanian Ministry of Education and Research [contract CEEX No. 60/2006]. The initial goal of the project was to generate a basic information technology [IT] infrastructure for collecting data on IBD nationwide. The IBDPROSPECT subsequently came under the remit of the Romanian Crohn’s and Colitis Club, a non-profit professional organization established as a professional and scientific body of the Romanian Society of Gastroenterology and Hepatology, and was interested in deepening the fundamental knowledge, diagnosis, and treatment of IBD in Romania.

The database can generate important insights into the dynamics of IBD in our country. The database allows the generation of new sets of specific variables for future projects that would have as their starting points the previously recorded data sets. Thus, IBDPROSPECT is a dynamic structure, among the few initiatives of its kind in Eastern Europe. The local ethics committees of all participating centres approved the study protocol and all participants provided signed informed consent before taking part.

The questionnaires include demographic information and clinical and laboratory data. These questionnaires were completed by the patient [questionnaires related to demographic information and symptoms] and by the physician [laboratory data, classification, complications, surgery for IBD-related complications]. This study involved IBD patients included prospectively in IBDPROSPECT from June 2006 to June 2014.

Although not stated in the IBDPROSPECT, in Romania patients can contact the family physician or call directly to a specialist gastroenterologist. In our study, the time between onset of symptoms and diagnosis is not divided into two intervals, as in the study of Vavricka et al.⁶ Only patients enrolled in university medical centres were included in this current study.

2.2. Methods

Diagnostic delay was defined as the time in months from the onset of symptoms attributable to subsequent IBD and diagnosis. In the first part of the article, we evaluate the correlation between diagnostic delay and patient- and disease-associated factors. We performed univariate analysis, followed by multivariate logistic regression. Factors that were analysed in relation to diagnostic delay were: sex [male versus female]; age at diagnosis [< 40 versus ≥40 years]; provenance [rural versus urban]; geographical region [Transylvania, Banat, Muntenia, Moldova: Figure 1]; ethnic group [Romanian versus other]; disease location [for CD: terminal ileum versus colon; for ulcerative colitis [UC]: proctitis versus more extended colitis [left-sided and extensive colitis]; family history of IBD; smoking at diagnosis; presence of extraintestinal manifestations; onset of symptoms during summer months; medical history of autoimmune diseases and/or diabetes mellitus; and history of appendectomy.

Figure 1.

The geographical regions of Romania.

The following extraintestinal manifestations were noted: ankylosing spondylitis; peripheral arthritis; uveitis; primary sclerosing cholangitis; pyoderma gangrenosum; erythema nodosum; and oxalate renal lithiasis. Patients with at least one extraintestinal manifestation were considered extraintestinal manifestation-positive. Disease location was assessed according to the Montreal classification.¹³ Regarding the ethnicity of patients, we considered two groups: Romanian; and all others [Hungarian, Roma, Bulgarian, German, Ukrainian, Tatar, Turkish, and Serbian]. Concerning the onset of symptoms in the summer months, we included in this group those patients who started to have symptoms during June, July, or August.

Among the possible factors associated with diagnostic delay, we considered a history of diabetes [associated with diabetic neuropathy and diabetic diarrhoea as a possible explanation of chronic diarrhoeal syndrome] and autoimmune diseases [with the possible shorter time to diagnosis due to increased suspicion of another autoimmune disease].

In the second part of the article, we referred only to the group with CD. Multivariate logistic regression was used to establish whether diagnostic delay affected: complications [intestinal stenoses, internal fistulas]; perianal disease [perianal fistulae]; surgery for any condition related to IBD; intestinal surgery [intestinal resection, colectomy, ileo- or colostomy]; and perianal surgery [perianal abscess drainage, fistulectomy, or fistulotomy]. Outcomes were analysed only at the time-point of first entry to IBDPROSPECT. Besides diagnostic delay, we analysed the impact of the following factors on the outcome described above: sex [male versus female; age at diagnosis [< 40 versus ≥ 40 years]; provenance [rural versus urban]; ethnic group [Romanian versus other]; disease location [terminal ileum versus colon]; family history of IBD; smoking at diagnosis; disease duration since diagnosis; and history of appendectomy.

2.3. Statistical analysis

Data were retrieved from the IBDPROSPECT database. Statistical analyses were performed using Statistical Package for the Social Sciences [release 20.0 for Windows; SPSS Inc., Chicago, IL] and R [v3.0.1; R Core Team, Vienna, Austria]. Quantitative data are presented as median, interquartile range, and minimum–maximum range; categorical data are presented as the percentage of the total group. Differences between groups were evaluated using a Wilcoxon unpaired test for quantitative data and a χ² test for categorical data. Independent variables with a p-value of less than 0.2 at univariate logistic regression were subsequently entered in a multivariate logistic regression analysis.

3. Results

3.1. Demographic and clinical data of the study population

The clinical characteristics of the study population are listed in Table 1. A total of 1518 patients were included in the IBDPROSPECT database. Complete information regarding diagnostic delay was available in 1196 patients. The median disease duration from diagnosis to enrolment was 4 months for the CD, UC, and indeterminate colitis [IC] groups alike.

Table 1.

Clinical characteristics of the study population.

	CD number [%]/ median, IQR, range	UC number [%]/ median, IQR, range	IC number [%]/ median, IQR, range
Number of patients	478	682	36
Sex
Female	249 [52.1%]	309 [45.3%]	17 [47.2%]
Male	229 [47.9%]	373 [54.7%]	19 [52.8%]
Ethnic group
Romanian	405 [84.7%]	588 [86.2%]	31 [86.1%]
Other	73 [15.3%]	94 [13.8%]	5 [13.9%]
Disease location [CD]
L1	199 [41.6%]
L2	85 [17.8%]
L3	185 [38.7%]
L4	16 [3.3%]
Disease location [UC]
E1		138 [20.2%]
E2		347 [50.9%]
E3		197 [28.9%]
Provenance
Rural	109 [22.8%]	202 [29.6%]	9 [25.0%]
Urban	369 [77.2%]	480 [70.4%]	27 [75.0%]
Region
Banat	52 [10.9%]	47 [6.9%]
Moldova	43 [9%]	119 [17.4%]
Muntenia	318 [66.5%]	233 [34.2%]	2 [5.6%]
Transylvania	65 [13.6%]	283 [41.5%]	20 [55.6%]
Age at enrolment	38, 28–52, 6–80	46, 35–59, 6–89	42.5, 32-57.5, 7-75
Age at diagnosis	33, 24–44, 6–77	38, 29–51, 4–82	37, 27.5-48, 5-69
Over 40 years old	157 [32.8%]	322 [47.2%]	16 [44.4%]
Under 40 years old	321 [67.2%]	360 [52.8%]	20 [55.6%]
Age at first symptoms	31, 23–43, 6–77	38, 28–51, 0–82	36.5, 25.75–47.25, 3–69
Disease duration since diagnosis [years]	4, 2–7, 0–40	4, 3–8, 0–51	4, 2–6.25, 0–19
Disease duration since first symptoms [years]	5, 2–9, 0–42	5, 3–8, 0–52	4.5, 2–7.25, 0–19

CD, Crohn’s disease; UC, ulcerative colitis; IC, indeterminate colitis; IQR, interquartile range.

3.2. Diagnostic delay and patient stratification

The diagnostic delay differed significantly between the CD and UC patient groups [p < 0.001] and between the IC and UC groups [p < 0.01]; the difference was not significant between the CD and IC groups [p = 0.51]. Of patients with CD, 50% were diagnosed within 5 months after symptom onset compared with only 1 month in UC and 4 months in IC patients. Moreover, 75% of patients with CD were diagnosed within 18 months after symptom onset compared with only 3 months in UC and 10 months in IC patients. The distributions of delay times are skewed toward short values in all groups, and the differences between groups are mostly derived from patients with long to very long delays. Instead of a submedian–supramedian division, we opted for thresholds at the 75% percentile of delay distributions, computed separately for the CD and UC groups. Each group was divided into subgroups of ‘long diagnostic delay’ [one-quarter of the patients] and ‘short diagnostic delay’ [the remaining three-quarters], similar to the method used by Vavricka et al. ⁶ The identified thresholds, which were used in all further analyses, were 18 months for CD and 3 months for UC. Due to the small number of patients with IC, this class of patients was excluded from regression studies.

3.3. Impact of disease-associated items on diagnostic delay

In univariate logistic regression analysis, location of the disease in the terminal ileum [L1] correlated with a long diagnostic delay in the CD group [p = 0.002]; being an active smoker at diagnosis and symptom onset during summer [as opposed to other times of year] were both risk factors for a long delay [p = 0.05 and P < 0.001, respectively]. All three variables remained significant in multivariate logistic regression analysis, identifying them as independent predictors of a long diagnostic delay [Table 2]. In the UC group, univariate logistic regression identified being over 40 years old at diagnosis as the only significant predictive factor for a long diagnostic delay [p = 0.04], with a protective effect; it retained significance after inclusion in multivariate logistic regression [Table 3].

Table 2.

Diagnostic delays [months] in patients with CD [n = 478], UC [n = 682], and IC [n = 36].

	Diagnosis	Univariable analysis			Multivariable analysis
		OR	95% CI	p-Value	OR	95% CI	p-Value
Sex, male/female
	UC	1.21	0.85–1.71	0.29
	CD	1.07	0.71–1.62	0.75
Ethnic group, Romanian/other
	UC	0.93	0.54–1.52	0.77
	CD	1.35	0.77–2.32	0.28
Age at diagnosis, under/over 40 years
	UC	0.7	0.49–0.99	0.04	0.68	0.48–0.97	0.04
	CD	0.93	0.59–1.44	0.75
Disease location
Proctitis versus extended [UC]	UC	1	0.64–1.53	1
L1 versus [L2 + L3] [CD]	CD	0.54	0.33–0.91	0.02	0.5	0.27–0.95	0.03
L1 versus L2 [CD]	CD	0.46	0.26–0.81	0.007
Perianal disease, absent/present	CD	1.44	0.782.6	0.23
EIM, absent/present
	UC	1.28	0.72–2.2	0.38
	CD	1.14	0.69–1.85	0.59
Past history of appendectomy, absent/present
	UC	1.64	0.77–3.32	0.18	1.78	0.83–3.63	0.12
	CD	1.33	0.79–2.2	0.27
Provenance, urban/rural
	UC	1.17	0.8–1.7	0.47
	CD	0.65	0.38–1.09	0.14	0.68	0.34–1.3	0.26
Smoking status
Non-smoker/smoker
	UC	0.99	0.62–1.54	0.95
	CD	1.58	1–2.48	0.05	2.09	1.19–3.68	0.01
Non-smoker/unknown
	UC	0.75	0.35–1.48	0.43
	CD	1.46	0.7–2.89	0.29	2.1	0.8–5.23	0.12
Onset of symptoms in summer, no/yes
	UC	1.19	0.78–1.78	0.41
	CD	3.32	2.01–5.53	< 0.001	3.35	1.99–5.69	0.000
Family history of IBD, negative/positive
	UC	0.38	0.06–1.34	0.19	0.38	0.06–1.38	0.21
	CD	0.68	0.15–2.16	0.55
Region
Transylvania/Banat
	UC	1.06	0.5–2.11	0.87
	CD	0.51	0.2–1.21	0.14	0.64	0.23–1.76	0.39
Transylvania/Moldova
	UC	0.91	0.54–1.5	0.72
	CD	0.47	0.17–1.2	0.13	0.6	0.2–1.72	0.35
Transylvania/ Muntenia	UC	1.08	0.72–1.6	0.72
	CD	0.88	0.5–1.62	0.68	1.04	0.51–2.2	0.92
History of autoimmune diseases, no/yes
	UC	1.02	0.22–3.47	0.98
	CD	1.52	0.56–3.75	0.38
History of diabetes mellitus, no/yes
	UC	0.6	0.17–1.62	0.36
	CD	0.81	0.18–2.64	0.75

Bolded binary outcomes denote the reference values for the odds ratio.

CD, Crohn’s disease; UC, ulcerative colitis; EIM, extraintestinal manifestations; IBD, inflammatory bowel disease; OR, odds ratio; CI, confidence interval.

Table 3.

Multivariate logistic regression analysis evaluating the correlation between a long diagnostic delay and disease-associated factors in UC and CD.

	OR	Stenosis [20.1%]			Intestinal fistula [7.7%]	p-Value
		95% CI	p-Value	OR	95% CI
Male	1 [Ref]	-	-	1 [Ref]	-	-
Female	0.86	0.53–1.4	0.55	0.74	0.36–1.53	0.42
Ethnic group
Romanian	1 [Ref]	-	-	1 [Ref]	-	-
Other	1.4	0.73–2.59	0.29	0.63	0.18–1.78	0.43
Age at diagnosis
Under 40 years old	1 [Ref]	-	-	1 [Ref]	-	-
Over 40 years old	0.75	0.44–1.26	0.28	0.29	0.09–0.71	0.01
Disease location
Ileal involvement	1 [Ref]	-	-	1 [Ref]	-	-
Colonic involvement	0.68	0.38–1.24	0.2	1.5	0.61–4.14	0.41
Provenance
Urban	1 [Ref]	-	-	1 [Ref]	-	-
Rural	1.09	0.61–1.92	0.76	0.8	0.28–1.95	0.64
Smoking status at diagnosis
Non-smoker	1 [Ref]	-	-	1 [Ref]	-	-
Smoker	0.59	0.34–1.01	0.06	0.57	0.25–1.23	0.17
Unknown	0.37	0.12–0.94	0.06	-	-	-
Family history of IBD
Negative	1 [Ref]	-	-	1 [Ref]	-	-
Positive	1.36	0.36–4.21	0.62	1.92	0.28–8.1	0.42
Appendectomy
No	1 [Ref]	-	-	1 [Ref]	-	-
Yes	2.12	1.21–3.66	0.01	4.1	1.91–8.78	0.000
Disease duration since diagnosis	1.04	1–1.08	0.04	1.01	0.95–1.07	0.77
Delay
< 25%	1 [Ref]	-	-	1 [Ref]	-	-
25%–50%	2.22	1.08–4.74	0.03	1.37	0.46–4.38	0.58
50%–75%	1.91	0.89–4.23	0.1	1.6	0.51–5.32	0.42
> 75%	3.38	1.66–7.2	< 0.01	2.64	0.94–8.24	0.08
	OR	Perianal fistula [9.2%]	p-Value	OR	Any surgery [26.1%]	p-Value
		95% CI			95% CI
Male	1 [Ref]	-	-	1 [Ref]	-	-
Female	0.54	0.27–1.04	0.07	0.63	0.41–0.97	0.04
Ethnic group
Romanian	1 [Ref]	-	-	1 [Ref]	-	-
Other	2.13	0.98–4.45	0.05	1.57	0.89–2.72	0.12
Age at diagnosis
Under 40 years old	1 [Ref]	-	-	1 [Ref]	-	-
Over 40 years old	0.78	0.36–1.58	0.5	0.86	0.54–1.37	0.53
Disease location
Ileal involvement	1 [Ref]	-	-	1 [Ref]	-	-
Colonic involvement	1.11	0.48–2.91	0.82	1.42	0.8–2.61	0.24
Provenance
Urban	1 [Ref]	-	-	1 [Ref]	-	-
Rural	1.39	0.64–2.86	0.38	1.52	0.92–2.48	0.1
Smoking status at diagnosis
Non-smoker	1 [Ref]	-	-	1 [Ref]	-	-
Smoker	1.06	0.5–2.19	0.87	1.26	0.78–2	0.34
Unknown	1.93	0.69–4.89	0.18	1.17	0.55–2.38	0.68
Family history of IBD
Negative	1 [Ref]	-	-	1 [Ref]	-	-
Positive	0.67	0.03–3.87	0.72	0.99	0.26–3.01	0.98
Appendectomy
No	1 [Ref]	-	-	1 [Ref]	-	-
Yes	0.31	0.07–0.91	0.06	2.23	1.34–3.72	< 0.001
	OR	Perianal fistula [9.2%]	p-Value	OR	Any surgery [26.1%]	p-Value
		95% CI			95% CI
Disease duration since diagnosis	0.99	0.92–1.05	0.65	1.04	1.01–1.08	0.02
Delay
< 25%	1 [Ref]	-	-	1 [Ref]	-	-
25%–50%	4.32	1.61–13.79	0.01	1.16	0.62–2.16	0.65
50%–75%	2.92	0.96–10.02	0.07	1.86	0.99–3.51	0.05
> 75%	2.9	0.98–9.82	0.07	1.96	1.07–3.65	0.03
	OR	Perianal surgery [9.8%]	p-Value	OR	Intestinal surgery [ 16.3%]	p-Value
		95% CI			95% CI
Male	1 [Ref]	-	-	1 [Ref]	-	-
Female	0.78	0.41–1.45	0.43	0.71	0.42–1.2	0.21
Ethnic group
Romanian	1 [Ref]	-	-	1 [Ref]	-	-
Other	0.92	0.36–2.09	0.86	1.82	0.94–3.41	0.07
Age at diagnosis
Under 40 years old	1 [Ref]	-	-	1 [Ref]	-	-
Over 40 years old	0.72	0.35–1.41	0.35	1.27	0.73–2.19	0.39
Disease location
Ileal involvement	1 [Ref]	-	-	1 [Ref]	-	-
Colonic involvement	1.44	0.63–3.72	0.42	1.14	0.58–2.4	0.71
Provenance
Urban	1 [Ref]	-	-	1 [Ref]	-	-
Rural	1.63	0.8–3.21	0.16	1.2	0.64–2.18	0.55
Smoking status at diagnosis
Non-smoker	1 [Ref]	-	-	1 [Ref]	-	-
Smoker	1.45	0.73–2.83	0.28	0.87	0.47–1.55	0.64
Unknown	1.49	0.47–3.97	0.46	1.28	0.53–2.85	0.57
Family history of IBD
Negative	1 [Ref]	-	-	1 [Ref]	-	-
Positive	1.41	0.21–5.53	0.66	0.8	0.12–3.06	0.77
Appendectomy
No	1 [Ref]	-	-	1 [Ref]	-	-
Yes	1.02	0.44–2.17	0.97	2.6	1.43–4.68	< 0.001
Disease duration since diagnosis	0.97	0.9–1.03	0.32	1.07	1.03–1.12	< 0.001
Delay
< 25%	1 [Ref]	-	-	1 [Ref]	-	-
25%–50%	1.77	0.73–4.6	0.22	1.22	0.56–2.69	0.62
50%–75%	1.33	0.48–3.78	0.58	2.24	1.06–4.9	0.04
> 75%	2.2	0.89–5.82	0.1	1.59	0.74–3.48	0.24

CD, Crohn’s disease; UC, ulcerative colitis; IBD, inflammatory bowel disease; OR, odds ratio; CI, confidence interval.

3.4. Impact of diagnosis delay on CD-related complications and surgery

In the second part of the study, which exclusively included CD patients, we evaluated demographic and clinical factors as predictors of six outcome measures specific to CD: intestinal stenosis; intestinal fistula; perianal fistula; surgery for any CD-related indication; surgery for perianal pathology; and surgery [resection] for intestinal pathology. Multivariate logistic regression was used separately for each outcome [Table3]. Diagnostic delay [used as one of the predictors] was categorized into the following four subgroups, along the quartiles of its distribution [similar to Schoepfer et al. ¹⁰]: < 25%, 0–1 months; 25%–50%, 2–5 months; 50%–75%, 6–18 months; > 75%, over 18 months.

Diagnostic delay was significantly and positively associated with intestinal stenosis (odds ratio [OR], 3.38; p < 0.001 for the delay subgroup >75%), along with a history of appendectomy [OR, 2.12; p = 0.01] and disease duration [OR, 1.04; p = 0.04]. For intestinal fistula, diagnostic delay was a significant predictor [OR, 2.64; p = 0.008 for the delay subgroup >75%], as well as a past history of appendectomy [OR, 4.1; p = 0.000]. Being over 40 years old at time of diagnosis was a significant protective factor for intestinal fistula [OR, 0.29; p = 0.01].

With respect to perianal fistula, belonging to an ethnic group other than Romanian [OR, 2.13; p = 0.05] was a risk factor; a diagnostic delay between 25% and 50% also correlated with this outcome [OR, 4.32; pP = 0.01]. Disease duration [OR, 1.04; p = 0.02], past appendectomy [OR, 2.23; p < 0.001], and belonging to diagnostic delay subgroups 50–75% [OR, 1.86; p = 0.05] and > 75% [OR, 1.96; p = 0.03] were all associated with a higher risk of surgery for any CD-related indication. Female sex [OR, 0.63; p = 0.04] was a protective factor. Perianal surgery, or any of the other studied variables, was not associated with diagnostic delay.

Intestinal surgery was predicted by a history of appendectomy [OR, 2.6; p < 0.001], disease duration [OR, 1.07; p < 0.001], and diagnostic delay subgroup 50–75% [OR, 2.24; p = 0.04]. An ethnicity different from Romanian displayed a risk trend [OR, 1.82; p = 0.07].

4. Discussion

Knowledge of IBD in the general population is poor, with only 32% and 21% of interviewees in the study of Angelberger et al.¹⁴ being aware of CD and UC, respectively. The unawareness of IBD symptoms among the general population and general practitioners can lead to a late presentation to specialist gastroenterologists. It seems that older people, those with a higher level of education, and those in urban areas have more information about these diseases.

To our knowledge, only a few studies in the literature have evaluated the risk factors for delayed diagnosis in IBD and the immediate impact of the delay on the disease course.^6,7 Symptoms occurring long before a diagnosis of IBD is established might be the initial IBD symptoms or might be completely unrelated to the IBD. There are several explanations for the diagnostic delay. In most cases, it is likely that low-intensity symptoms are not deemed worthy of further exploration or are investigated with routine screenings that might miss the IBD diagnosis.¹⁵

Our current results agree with those of previous studies, which have noted a much longer diagnostic delay for CD than for UC. This difference might be because CD symptoms are much more variable than UC symptoms. Atypical features such as fatigue, chronic diarrhoea, abdominal pain, fever, and weight loss are the most prominent symptoms of CD whereas, in the case of UC, rectal bleeding is often the unsettling symptom that leads to early referral and timely diagnosis.

Whereas there is no doubt that an UC diagnosis is reached in less time than a CD diagnosis, the existence of a difference in IBD diagnostic delays between younger and older patients is still a matter of debate. Some studies have shown a greater diagnostic delay in young patients,⁶ whereas others have shown the opposite.⁴ In the case of older patients, there are several explanations for the prolonged duration of symptoms before a final IBD diagnosis is reached. On one hand, the differential diagnosis of chronic abdominal pain and chronic diarrhoea syndromes is much broader in older patients. On the other hand, the attitude of medical practitioners and health insurers toward an older patient might be different toward a younger one.¹⁷ Finally, the pathogenesis of IBD might vary substantially with age.¹⁸

From the other perspective, symptoms in younger patients may be more frequently interpreted as belonging to irritable bowel syndrome.⁶ In addition, rectal bleeding is often considered to be caused by a benign perianal pathology [haemorrhoids, anal fissures]. In our current study, as in that of Vavricka et al.,⁶ young age at diagnosis was found to be a risk factor for a long diagnostic delay.

Some studies in the literature have investigated the seasonal variation in the onset of IBD symptoms.^19,20 Some reports reveal no seasonal differences,¹⁹ whereas others have shown that onset is more common during the spring and summer months than in autumn and winter.²⁰ In our present study, we investigated whether the onset of symptoms during the summer months correlated with diagnostic delay. If the onset of CD symptoms was during summer, there was a greater possibility of a long diagnostic delay. At this time of year, digestive infections are more frequent and that may explain—at least for a short time—diarrhoeal syndrome, leading to a delay in CD diagnosis. Another explanation may be a seasonal variation in the immune response; in this case, a gastrointestinal infection may be responsible for triggering IBD onset.^20,21

Regarding the influence of CD phenotype, a location in the small intestine is also associated with a long diagnostic delay. These results are similar to those obtained by Vavricka et al. ⁶ The most plausible explanation for this finding is that, in Romania, investigation of the small intestine occurs later in the investigation of patients with CD.

Regarding the link between smoking and diagnostic delay, this can be seen from two perspectives, at least when it comes to patients with CD. On the one hand, smokers should have a shorter interval between the onset of symptoms and time of diagnosis, due to a more aggressive disease.¹¹ On the other hand, smokers are most often part of disadvantaged groups, without health insurance and with less access to medical services.²² In our present study, smoking at diagnosis was associated with a long diagnostic delay.

We hypothesized that a medical history of diabetes would be associated with a longer diagnostic delay, due to the possible association with diabetic neuropathy and diabetic diarrhoea.²³ Chronic diarrhoea could be interpreted in this context, leading to a long diagnostic delay. Regarding the association with autoimmune diseases, the high degree of suspicion for a second autoimmune pathology could shorten the time to diagnosis. In our current investigation, none of these two factors was associated with a diagnostic delay.

Many studies, even those of referral centre-based cohorts, have shown that CD displays an inflammatory pattern at onset, without complications, becoming penetrating/stenosing over many years when complications are present.²⁴ Cosnes et al.²⁵ showed that over a 20-year disease course, the proportions of cases with inflammatory, stenosing, and penetrating disease were 12%, 18%, and 70%, respectively, and that after 20 years of evolution most patients with CD will undergo surgery at least once. Approximately 30% of patients with CD are operated on in the first year after diagnosis. The surgery rate after the first year is 5% per year, so the cumulative risk of surgery in CD is 50–80%, with half of these patients requiring re-intervention during the disease course.²⁶

In our current analyses, diagnostic delay was found to be an independent predictor of complications such as stenosis and intestinal fistulas, surgery requirement, and the risk of intestinal resection in CD, similar to the results of Schoepfer et al.¹⁰ Disease duration was also associated with the need for surgery for CD-related complications. An ethnicity other than Romanian [Hungarian, Roma, Bulgarian, German, Ukrainian, Tatar, Turkish, and Serbian] was significantly associated with perianal disease in BC, but due to the small number of patients in this group, the results cannot be interpreted definitively, warranting further separate studies on each ethnic group.

4.1. Strengths/limitations of the study

The strength of our present study lies in the fact that it is the first report from an Eastern European country to evaluate the diagnostic delay in IBD, the risk factors for this delay, and the impact it has on the CD outcomes, complications, and surgery for CD-related complications. There are also several limitations to our study, however. One limitation is that symptoms at IBD onset are not stipulated in the IBDPROSPECT registry. Another is that the cohort of patients we included is only partially population based because the patients were recruited only from university medical centres. In Eastern European countries, as well as in Romania, with few exceptions the role of private practice is still limited; most patients are treated in the public health system.¹ Our study also could not specify data such as education level, non-steroidal anti-inflammatory drug intake, or oral contraceptive use. Lack of consideration of treatments in the multivariate analysis of the risk factors for the complications of Crohn’s disease and the need of surgery may be deemed a major limitation. Even if the study had been conducted prospectively and was based on a national registry, a confounding factor may still be recall bias regarding the timing of onset of IBD symptoms.

5. Conclusions

In Eastern European countries, there is a considerably long time between the onset of symptoms and the time of diagnosis in inflammatory bowel disease, longer for CD than for UC. In addition, ileal location, being a smoker at diagnosis, onset of symptoms in the summer months in CD, and a young age at diagnosis in UC, are predictors of a long diagnostic delay for this condition. A long diagnostic delay is a predictor of surgery in CD. In light of these findings, we believe that implementation of awareness programmes in the general population and among physicians, especially among general practitioners, is the next step for the management of IBD in Romania.

Funding

This communication is published under the framework of the European Social Fund, Human Resources Development Operational Programme 2007–2013 [project number POSDRU 159/1.5/138776], and is supported by internal grants from the Iuliu Hatieganu University of Medicine and Pharmacy awarded to RH, No. 1491 / 4 / 28.01.2014.

Conflict of Interest

All authors declare that they have no conflict on interest regarding this manuscript.

Author Contributions

RZ and AT mad emajor contributions in writing article, data analysis, and data interpretations. The others authors have made substantial contributions to the concept and design and acquisition of data, or analysis and interpretation of data.