Patient adherence rates with surveillance colonoscopy are not well documented. In the only study that directly addressed this issue, Woolrich et al. reported on 7 patients of their cohort of 121 that were found to have cancer [18]. In two of these patients, previous colonoscopy had found dysplasia in the setting of quiescent disease, and neither of these patients were adherent with recommendations for close follow up colonoscopy or colectomy. It was the conclusion of the authors that quiescent disease was a risk factor for non-adherence with physician recommendations.

What data do we have that would compel a patient to continue taking medication in the setting of well-being? A prospective 2-year follow up on a cohort of patients with quiescent UC was done to help answer this question [19]. Patients in remission were enrolled and then stratified by adherence based on the previous 6-month pharmacy refill data. At 6 months, 12 patients (12 %) had clinical recurrence of disease symptoms, all of whom were noncompliant with medication. At 12 months, 19 of 86 patients had recurrent disease, 13 (68 %) of whom were noncompliant. Multiple Cox proportional hazards model revealed that patients not compliant with medication had more than fivefold greater risk of recurrence than the compliant patients (hazard ratio 5.47, 95 % confidence interval 2.26–13.22, p < 0.001). As part of the study, non-adherent patients were asked why they were not taking their medications [20]. The majority stated that they simply forgot one of their doses. Fewer than 10 % of patients complained of side effects and cost.

There are now several studies that suggest that documented medication consumption is protective for colon cancer, which is an important concern for the long-term natural history of UC. Moody et al. studied 168 patients with UC diagnosed between 1972 and 1981 and correlated sulfasalazine non-adherence with risk of colorectal cancer [21]. A patient was classified as non-compliant if there was clear evidence in the medical record of medications not taken, or upon the advice of a physician medication was discontinued. Their crude colectomy rate was 23 % in 10 years with a 3 % rate in those patients on maintenance sulfasalazine, and 31 % in patients either non-compliant or off all medications. Since the authors found a colectomy rate and cancer incidence similar to previously published series, they concluded that medications were beneficial in reducing cancer risk. In a second retrospective case control study, Pinczowski found that a record of at least a 3-month history of therapy with sulfasalazine had a protective effect for colon cancer [22]. There was a 62 % reduction in risk with any history of therapy in the 102 patients studied. It is difficult to interpret this finding however, as documentation of dose and duration of therapy for each patient was imprecise. More recently, a published meta-analysis suggests that the risk reduction for dysplasia and colorectal cancer is significant [23].

Eaden and colleagues found in a case-control study that mesalamine at a dose of 1.2 g/day or greater reduced colorectal cancer risk by 91 % in patients with UC compared to no treatment [24]. There was also a protective effect of >2 visits to the physician per year, but the same was not found for the number of surveillance colonoscopies. More data on mesalamine comes from investigators at Mount Sinai Medical Center in New York who followed patients whose colon biopsies were indeterminant for dysplasia [25]. Those patients on 2 g or more of 5-ASA per day did not progress to definite dysplasia, suggesting that any chemoprotective effect occurs early in the cancer progression pathway. Rubin and colleagues studied the University of Chicago experience with all forms of 5-ASA in ulcerative colitis patients [26]. The use of at least 1.2 g daily of a 5-ASA carried a 76 % relative risk reduction for the development of colorectal cancer, when controlled for disease extent, duration and folic acid use. In contrast, IBD patients on 5-ASA agents followed in Canada did not appear to have the same protective effect [27].

There have also been recently published data on the effects of non-adherence on health care costs. Kane et al. demonstrated that those UC patients without an active prescription cost 30 % more in hospital and outpatient costs than patients who were taking medication [28]. In addition, those taking less than 80 % of medication were also associated with higher health costs. Yen et al. performed a cost effectiveness study and showed that over a year period adherence with 5-ASA was associated with a higher health care quality of life than those who were not [29].