Diverticular disease of the colon - risk factors and validation of diagnosis

Abstract

Diverticular Disease (DD) is a common gastrointestinal disease, particularly among the population in western countries. The aetiology of DD is complex and have not yet been established conclusively. Dietary habits and lifestyle choices such as, physical inactivity, obesity, and smoking, affects the development of the disease. The prevalence of DD is increasing with age, at the age of 65 years, 66% have DD. Inflammation (diverticulitis) is the most common complication with up to 5%, which may result in bowel perforation, obstruction, and/or bleeding. The aim of this thesis was to identify risk factors for DD, validate the quality of registered ICD-codes in the Swedish National Patient Register and compare rates of anastomotic leakage (AL) and other possible risk factors for AL between patients with planned surgery for DD versus colonic cancer.

In Paper I, the intake of oral/inhaled corticosteroids, indometacin or aspirin and if they influence the risk of diverticular disease were evaluated. A population-based prospective cohort of middle-aged women in the Swedish Mammography Cohort were assessed. Use of corticosteroids, indometacin or aspirin in 1997 was determined from questionnaires. Cases of diverticular disease were identified from the Swedish national registers until the end of 2010. The total cohort consisted of 36 586 middle-aged women, out of these, 674 (1.8 %) were admitted to hospital at least on one occasion with DD. A significant correlation was observed between corticosteroids, especially those inhaled, and hospitalization due to diverticular illness.

In Paper II, the association between dietary fibre intake and hospitalisation due to DD was investigated in a cohort study. The Swedish Mammography Cohort and the Cohort of Swedish Men were linked to the Swedish Inpatient Register and the Causes of Death Register. Data of dietary fibre consumption were gathered through questionnaires. The risk for hospitalisation was reduced by 30% in the women with the highest intake (median 12.6 g/day) of fruit and vegetable fibres compared to those with the lowest intake (4.1 g/day). The results were replicated in the male population, where the men with the highest intake (median 10.3 g/day) had a 32% lower risk of hospitalization than the men with the lowest intake (median 2.9 g/day).

In Paper III, the validity of the coding for DD in the Swedish National Patient Register was studied. The study included 323 and 327 patients from the years 2002 and 2010, respectively. The overall Positive Predictive Value for both years and all diagnoses was high, 95% (95% CI: 93-96). The Positive Predictive Value for the year 2010 was slightly higher 98% (95% CI: 95-99) than for the year 2002, 91% (95% CI: (87-94). The higher Positive Predictive Value for the year 2010 may be due to increased use of computed tomography in the diagnosis work.

In Paper IV, the rates of anastomotic leakage and potential risk factors for anastomotic leakage between patients undergoing surgery for diverticulosis versus colonic cancer was studied within an Enhanced Recovery After Surgery protocol. The study population consisted of patients operated on left sided colonic / sigmoid resection with diverticulosis and colonic cancer in Sweden between the years 2010-2020. There was no difference in anastomotic leakage between both diagnosis groups, colon cancer (4%) versus diverticulosis (3.8%) (p =0.992). When compared to laparoscopic surgery, there was more leakage among patients with diverticulitis who had open surgery. Anastomotic leakage was more common among cancer patients who had undergone laparoscopic surgery (adjusted OR 4.02, 95% CI (1.35- 11.91), p=0.01).