Table 4. Illustration of additional cases of cancer and IBD following presentation with each studied symptom above what would be expected by the 12-month general population incidence of either condition, as applied to the age and sex composition of our study population.
| Cancer | IBD | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Men | Women | Persons | |||||||
| Observed | Expected* | Additional | Observed | Expected* | Additional | Observed | Expected* | Additional | |
| Abdominal bloating/distension | 450 | 402 | 48 | 982 | 692 | 290 | 1,029 | 28.7 | 1,000 |
| Per 1,000 patients** | 16.5 | 14.7 | 1.8 | 13.3 | 9.4 | 3.9 | 10.2 | 0.3 | 9.9 |
| Abdominal pain | 5,841 | 4,282 | 1,559 | 6,749 | 5,082 | 1,667 | 10,386 | 257 | 10,129.3 |
| Per 1,000 patients** | 17.7 | 13.0 | 4.7 | 20.9 | 15.7 | 5.2 | 12 | 0.3 | 11.4 |
| Change in bowel habit | 2,101 | 879 | 1,222 | 1,471 | 838 | 633 | 2,859 | 29.5 | 2,829.5 |
| Per 1,000 patients** | 46.4 | 19.4 | 27.0 | 34.1 | 19.4 | 14.7 | 27 | 0.3 | 26.5 |
| Dyspepsia | 3,072 | 3,005 | 67 | 3,091 | 3,256 | −165 | 4,939 | 146 | 4,792.5 |
| Per 1,000 patients** | 14.1 | 13.8 | 0.3 | 14.4 | 15.2 | −0.8 | 10 | 0.3 | 9.3 |
| Dysphagia | 1,601 | 737 | 864 | 1,041 | 696 | 345 | 723 | 23.9 | 699.1 |
| Per 1,000 patients** | 42.8 | 19.7 | 23.1 | 29.1 | 19.4 | 9.6 | 8 | 0.3 | 8.1 |
| Rectal bleeding | 3,730 | 1,567 | 2,163 | 2,911 | 1,341 | 1,570 | 6,351 | 66.8 | 6,284.2 |
| Per 1,000 patients** | 32.0 | 13.4 | 18.5 | 25.8 | 11.9 | 13.9 | 27 | 0.3 | 26.7 |
As an example, 43 out of 1,000 men presenting with dysphagia would be diagnosed with cancer in the following year given the observed findings (see super-row 5 column 2, corresponding to PPV of 4.28% in Table 2). The corresponding number of cancer cases expected by applying the general age-specific population incidence would have been 20 (column 3). Therefore, dysphagia presentation is associated with 23 additional cases of cancer per 1,000 patients (column 4).
Across the symptom cohorts, it can be seen that the number of additional cancer cases is low for symptoms that are both associated with lower risk and younger median age at presentation (e.g., abdominal bloating/distension and dyspepsia, which have relatively lower PPVs for cancer (see Table 2) and lower median age at presentation in our study population of 53 years and 56, respectively; see Table 1). Conversely, the number of additional cases is greater for symptoms associated with relatively higher cancer risk and higher median age at presentation (e.g., change in bowel habit and dysphagia, which have relatively higher cancer risk; see Table 2 and median age at presentation of 63 years for both; see Table 1).
* Using age-specific population-based incidence estimates applied to the age (age/sex) structure of each of the 6 symptom cohorts.
** Per 1,000 patients presenting with the relevant symptom.
IBD, inflammatory bowel disease; PPV, positive predictive value.