Short interval between symptom onset and medical care as an indication of rapid onset of interstitial cystitis/painful bladder syndrome

Authors


John W. Warren, University of Maryland School of Medicine, 10 South Pine Street, Room 9–00 MSTF, Baltimore, MD 21201, USA. e-mail: jwarren@medicine.umaryland.edu

Abstract

OBJECTIVE

To assess how many patients with interstitial cystitis/painful bladder syndrome (IC/PBlS) obtain medical care soon after symptom onset, and to determine how these patients differ from those who have medical care later.

PATIENTS AND METHODS

In a risk-factor study of IC/PBlS, we recruited women with IC/PBlS symptoms of ≤12 months’ duration and carefully identified the onset date, first medical encounter and early clinical findings.

RESULTS

Of 138 women with IC/PBlS for ≤12 months, 49% sought medical care within 7 days and another 31% within 4 weeks. These patients had no easier access to medical care but rather experienced symptoms differently, with more evidence for discomfort and inflammation. However, subsequently they did not differ from those with more gradual onset in features indicative of IC/PBlS, i.e. Hunner’s ulcers, petechial haemorrhages, symptoms, medications, or quality of life.

CONCLUSIONS

A shorter time to the first medical encounter was not a function of greater access to medical care but instead was associated with more discomfort and evidence of inflammation, as distinct from those with more gradual onset. The 6-month follow-up of both groups indicated that no significant differences in IC/PBlS symptoms persisted. A careful study of rapid-onset IC/PBlS might provide clues to the pathogenesis of this enigmatic disease.

Abbreviations
IC/PBlS

interstitial cystitis/painful bladder syndrome

EPIC

Events Preceding Interstitial Cystitis.

INTRODUCTION

Interstitial cystitis/painful bladder syndrome (IC/PBlS) is a chronic disease and urologists (the specialists who have most often studied it) generally are not referred these patients until months or years after the onset of symptoms. Consequently, reports of clinical features at the onset of IC/PBlS are few. We conducted a case-control study to identify risk factors for IC/PBlS and, to maximize the recollection of exposure before its onset, we sought women who had had IC/PBlS symptoms for ≤12 months. Consequently, through interview and medical records, we carefully established the date of symptom onset and the early longitudinal history of IC/PBlS. Many patients sought and received medical care soon after the onset of symptoms, and here we report how these patients differed from those obtaining medical care later.

PATIENTS AND METHODS

Events Preceding Interstitial Cystitis (EPIC) was a case-control study intended to identify risk factors for IC/PBlS. Recruitment was throughout the 48 contiguous States and methods included website links, brochures, posters, support group and urology meetings, letters to urologists, and emails to patient-support groups. We received responses primarily by e-mail and toll-free telephone calls. To enhance the recall of antecedent exposures, we recruited patients with IC/PBlS symptoms of ≤12 months. To avoid confounding by prostate disease, only women were eligible. Enrolment criteria were: ≥4 weeks of a syndrome of perceived bladder pain (≥3/10 on a Likert scale) and two of urinary frequency (≥8 times/24 h), urgency (≥3/10) or nocturia, and access to pertinent medical records. Exclusion criteria were 12 diseases mimicking IC/PBlS symptoms, as established by the National Institute of Diabetes, Digestive and Kidney Diseases [1] or, to avoid confounding by neurogenic bladder, a history of selected neurological diseases (spinal cord injury, stroke, Parkinson’s disease, multiple sclerosis, or spina bifida). Initial UTI was not an exclusion criterion if symptoms persisted with negative urine analysis and/or negative urine culture. As we were seeking individuals as soon as possible after the onset of symptoms, we required no diagnostic urological procedure.

Participants identified medical practitioners who provided care from 12 months before the onset of symptoms through any subsequent urological procedures, and signed releases to allow us access to the corresponding medical records. For these 138 participants, we received 455 (95%) of the requested 480 medical records, including 126 (91%) of those of the first medical encounter after the onset of symptoms. The medical records were abstracted by a physician (J.W.W.) to confirm the diagnosis of IC/PBlS, exclude other possible confounding diseases, and note clinical findings at the onset of IC/PBlS and in its early longitudinal history.

Determining the onset date of IC/PBlS symptoms was a five-step process: (i) the patient’s estimate; (ii) probing questions about earlier bladder pain, frequency or urgency; (iii) probing questions about similar episodes lasting ≥4 weeks in the previous 5 years; (iv) medical record confirmation; and (v) the patient’s final concurrence. Medical records confirmed onset dates in 133/138 (96%) of the patients; for the remaining five the discrepancies between medical records and patient recall were discussed with patients and an onset date agreed. Patients passing their medical record review then had telephone interviews at the baseline and every 6 months.

Patients were assigned to groups denoting the interval between the onset date and, as determined from the medical records, the first medical encounter. These groups were compared by 70 variables clustered in the following sets: demographics (four), access to medical care (six), possible inciting events (13), initial findings (10), medical evaluation (19), and longitudinal history (18).

Medical record and interview data were compared using Cochrane’s Q, and groups were compared by Pearson’s chi-square and one-way anova. Because of the many variables screened, we used a significance test of P ≤ 0.01 to specify differences between groups.

RESULTS

Of the initial 576 women screened for the EPIC study, the first 138 who met all inclusion and exclusion criteria are reported here; their characteristics have been described previously [2–4]. Briefly, median baseline symptoms, at their worst, were a pain score of 9 and urgency of 8; 119 patients (86%) had a urinary frequency of ≥11 times/24 h and 98 (71%) had nocturia of ≥3 times/night. The mean score on the IC Symptom Index [5], also known as the O’Leary-Sant Index, was 14.4, similar to that of IC/PBlS cases entering the large clinical trial in which this index was first used [6]. In all, 72 patients had had hydrodistension; six (8%) had a Hunner’s ulcer and 64 (89%) had petechial haemorrhages. Of 22 patients who had a potassium sensitivity test, 20 (91%) reported that the potassium solution elicited more pain or urgency than the initial water instillation. Of those studied by urodynamics, 17/18 (94%) reported a first sensation to void of <150 mL and 12/21 (57%) reported a maximum cystometric capacity of <350 mL. A urologist or gynaecologist wrote the specific diagnosis of IC/PBlS in the medical records of 119 (86%) patients. By the time of the baseline interview, 124 women had received one or more standard IC medication (oral pentosan polysulphate, amitriptyline, or hydroxyzine; or intravesical dimethylsulphoxide, heparin, or lidocaine). All 138 had received such medications and/or had a documented IC/PBlS diagnosis by a urologist or gynaecologist.

From the medical records we identified the date of the first medical encounter in 123/138 (89%) patients and for these we plotted the distribution of intervals between symptom onset and first medical encounter; Fig. 1 shows that most such intervals clustered within the first month of symptoms. Based on this plot we divided patients into three groups by interval, i.e. 0–7 days, 1–4 weeks and ≥4 weeks. For the 15 patients for whom we could not specify the exact date of the first medical encounter, medical records allowed a distribution of seven, three and five, respectively, into the three intervals. Thus, 67 (49%) sought medical care within 7 days, 43 (31%) at 1–4 weeks and 28 (20%) at >4 weeks after onset.

Figure 1.

Distribution of IC/PBlS cases by interval between symptom onset and first medical encounter.

We first hypothesized that the time to the first medical encounter was a function of access to medical care, and examined differences in medical insurance, income level, highest educational degree attained, and the pre-existence of 17 non-bladder diseases (a measure of a previous relationship with a physician). There were no significant differences in these variables among the three groups, and thus we inferred a similar access to medical care. Of interest was that 137/138 (99%) patients had medical insurance at the onset of IC/PBlS.

We then hypothesized that those who received medical care earlier had more severe disease; hence we compared the symptoms at the onset date (prevalence, mean number and combinations), features of the initial medical encounter (physician type, non-bladder symptoms, tender bladder or urethra, and features denoting UTI, including urine analysis, urine culture and antibiotic prescription), and the worst IC/PBlS symptoms experienced by the patients between onset and interview. Patients obtaining medical care more quickly reported significantly greater pain when symptoms were at their worst (Table 1). Also, significantly more patients who obtained medical care quickly had non-bladder symptoms (fever sensation, chills, sweats, or flank pain), were seen by emergency room physicians, had greater pyuria, and were prescribed antibiotics.

Table 1.  Measures of IC/PBlS severity in 138 patients
CharacteristicInterval, onset to medical encounterP
0–7 days1–4 weeks>4 weeks
  1. NS, not significant.

No. of patients674328 
Initial encounter, n (%):
Type of physician (emergency room)13 (19) 5 (12) 2 (7)0.002
Mean no. of:
Non-bladder symptoms 0.27 0.12 0.140.004
Urinary leukocytes (high power field or µL)18 8 60.001
Prescribed antibiotics, %8567480.003
Worst symptoms between onset and interview, n (%) or mean:
 Pain, mean (0–10) 8.6 8.5 8.20.002
 suprapubic49 (73)34 (79)22 (79)NS
 better after urination32 (48)24 (56) 11 (39)NS
 worse with filling51 (76)31 (72)19 (68)NS
 frequency, ≥11/24 h60 (90)36 (84)23 (82)NS
 Urgency, mean (0–10) 7.6 7.2 7.5NS
 Nocturia, ≥3/night47 (70)31 (72)20 (71)NS

The intervals between IC/PBlS symptom onset and first receipt of a standard IC/PBlS medication were significantly different, with means of 102, 125 and 170 days, respectively (P = 0.003). This was entirely due to the different intervals before the first medical encounter. There were no significant differences in intervals between the first medical encounter and the first receipt of an IC/PBlS medication (97, 100 and 82 days, P = 0.27).

We examined the three groups by additional features, including demographics, past medical history, subsequent evaluation, follow-up symptoms, and types of standard IC medications, but found no significant differences. Specifically, among these three groups, there were no differences in proportions with Hunner’s ulcers (8%, 10% and 8%), or petechial haemorrhages. At 6 months after the baseline interview there were no significant differences in any IC/PBlS symptom, mean total score on the IC Symptom Index, use of each IC/PBlS medication, or in quality-of-life measures. For all the comparisons above, we redistributed patients into two groups, with intervals of ≤4 and >4 weeks, and the findings were essentially the same.

DISCUSSION

Little has been published about early IC/PBlS; most clinical series describe IC/PBlS cases years after the onset of the disease. We think that the present study is the first to specifically recruit patients with recent onset, to carefully identify the onset date, to acquire medical records to confirm that date, to assess early clinical and laboratory findings, and to prospectively follow incident cases. Based on our interviews and medical record reviews, we are convinced that for these patients the date of onset of their IC/PBlS symptoms was a fulcrum; before that they had their routine health and after that they had a painful bladder disease.

After the onset of symptoms most of the patients sought medical care relatively quickly, 49% obtaining it within 7 days and another 31% at 1–4 weeks after onset. Those receiving care earlier did not do so because of easier access to medical care. Rather they appear to have experienced symptoms differently, with more evidence for discomfort and inflammation.

The rapid acquisition of medical care might be perceived as a surrogate for rapid onset of the disease. Indeed, rapid onset of IC/PBlS has been reported; Held et al.[7] calculated that 33% of 902 patients with established IC/PBlS noted that their ‘symptoms fully manifested within one month of onset’ Simon et al.[8] reported that of 424 patients with established IC/PBlS from several clinical practices, 40% stated that the onset of their symptoms was ‘sudden’. Two pioneering investigators also reported rapid onset IC/PBlS. Hand [9] noted that in 223 cases he was impressed with ‘the rapidity with which the symptoms can progress’ and Oravisto [10] noted that the disease ‘does not as a rule progress continually but reaches final stage rapidly’.

Although varying definitions make direct comparison difficult, the present study appears to have more patients whose IC/PBlS was of rapid onset than in the reports cited. There are several possible reasons for this. One is selection bias; to be eligible for the present study, patients had to know their onset date and to have pain as part of their syndrome, and, if episodic, symptoms had to persist for ≥4 consecutive weeks. Also, patients had only 12 months to experience the onset of disease, determine that it might be IC/PBlS, become aware of our study, and contact us. Patients with IC/PBlS and lacking one or more of these features would not have been enrolled in the study.

Another possibility is that these patients with rapid onset of bladder pain and urinary symptoms do not have IC/PBlS. However, in each group, similar proportions of patients had subsequent features characteristic of IC/PBlS, i.e. Hunner’s ulcers [1], petechial haemorrhages [1], documented IC/PBlS diagnosis, receipt of IC-specific medications [11,12], symptoms [1,5], and measures of quality of life, i.e. even though those with more rapid onset initially had somewhat more severe symptoms, these patients became indistinguishable from those who had a more gradual onset.

A final possibility is intriguing; that our findings are an accurate representation of recent onset, or incident, IC/PBlS, and that the earlier reports cited are likewise an accurate representation of established, or prevalent, IC/PBlS. For this, one would have to postulate a ‘survivor effect’, i.e. that rapid-onset cases would be more likely than those of gradual onset to enter long remissions or cures, and thus be under-represented in case series and surveys of prevalent cases.

It seems clear that a proportion of patients with IC/PBlS have a rapid onset of their disease. Many knowledgeable investigators postulate IC/PBlS as heterogeneous and perhaps several diseases [10–15]. Interestingly, the patients with a rapid onset of IC/PBlS might be successfully investigated for identifiable event(s) that initiated their disease. One might extrapolate from the pathophysiology that causes rapid onset of other diseases such as infection, trauma, obstruction, ischaemia, etc. Nevertheless, the present results suggest that if IC/PBlS comprises different inciting factors, or even different diseases, these various groups merge into a common clinical pathway, at least in the early months of symptomatic disease. Whether this similarity endures or whether disease trajectories will later diverge requires a longitudinal study of these incident IC/PBlS cases.

ACKNOWLEDGEMENTS

The study was funded by the National Institute of Diabetes, Digestive and Kidney Diseases (DK 064880) and was approved by the University of Maryland Institutional Review Board. We thank the women with IC/PBlS who are participating in this study, to their physicians who care for them, to the Interstitial Cystitis Association and the Interstitial Cystitis Network who assisted in recruitment, and to our colleagues, Vivian Brown, Suzanne Markowitz, Linda Horne, and Patricia Langenberg.

CONFLICT OF INTEREST

None declared.

EDITORIAL COMMENT

In this interesting paper the authors examined whether patients with IC/PBlS who rapidly seek medical care soon after symptom onset differ from those who delay in seeking care. They assessed the patients 6 months later and found no difference in the groups. The article raises some tantalizing questions.

Does the data suggest that early identification and treatment is ineffective in preventing symptomatic expression of the disease? If so, this is a major finding and would have clinical repercussions, as it would suggest that one need not and should not rush to diagnosis and institution of treatment in patients with mild to moderate symptoms.

Does symptom severity itself correlate with gradual or sudden onset? The authors determined that those who seek medical care early have worse symptoms than those who do not, but the onset of the symptoms over time is not well appreciated between the groups.

I wonder whether the key factor the authors are seeking might be the initial attempt to schedule an appointment with a provider, rather than the date seen. Many specialists have waiting periods for first appointments of >4 weeks, and the time to appointment might not reflect exactly what the authors are trying to ascertain as much as the time to first seeking medical care. Many offices tend to transfer these patients to their primary providers for a ‘UTI’, in favour of filling emergency slots with cancer patients. Others are not seen at all for several weeks, and the difference between an interval of 7 days and >4 weeks might not be a material one. That might be why the data look as they do, with no difference in the groups.

I think it is most important to compare sudden onset of severe symptoms with sudden or gradual onset of mild to moderate symptoms, and see how these patients fare at 6–12 months. Natural history and perhaps treatment would both be factors playing a role, and understanding these would yield extremely important data.

Philip Hanno,
Professor of Urology, Hospital of the
University of Pennsylvania, 9 Penn Tower,
3400 Spruce Street, Philadelphia,
Pennyslvania 19104, USA

Ancillary