Analysis and statistics
Patients were categorized into group I (investigations to confirm a working diagnosis were performed) and group NI (no investigations were performed). Three subgroups were identified within group I. We included all patients with a duodenal, gastric, or unspecified ulcer in group I-ULCER; group I-GERD included all patients with symptomatic or erosive gastroesophageal reflux disease (GERD); and group I-FUNCTIONAL included patients with only gastritis or with no imaging abnormalities. Patients with an ulcer and esophagitis were placed in group I-ULCER for their patient characteristics and medication prescription and in both groups I-ULCER and I-GERD for their medication indication, diagnostic tests, and eradication of H pylori.
Data were analyzed with the use of Statistical Package for the Social Sciences software (version 7.5.3). The chi-square test was used for comparison of proportions. Significance was set at = .05 (two sided).
Results
General characteristics and medication prescription
Of 46,813 patients listed with the 24 general practices, 988 (2.1%) were identified as long-term users of ASDs. Of these 988 patients, 66 were excluded because of ASD use for gastric or esophageal cancer, nongastric-related indications such as renal failure, or discontinuation of visits to the GP (patient moved or was a temporary visitor). The demographic and prescription characteristics of the remaining 922 patients are presented in Table 1. All patients had used H2-blockers and proton pump inhibitors for 12 weeks or more during the previous year.
Group I-ULCER consisted of 271 ulcer patients; group I-GERD, of 294 patients with reflux disease; and group I-FUNCTIONAL, of 127 patients with functional dyspepsia. Group NI consisted of 230 patients who did not undergo any confirmatory diagnostic testing (no endoscopy or barium study). Among the long-term users, treatment was more frequently prescribed for women than men (55% vs 45%, respectively; P < .05). Women were more likely not to undergo any diagnostic investigation (28% vs 21%, P < .05). If investigated, however, women were less likely than men have an ulcer (30% vs 50%, P < .05) and more likely to have functional dyspepsia (25% vs 11%, P < .05).
Overall, ranitidine was the drug most commonly prescribed. The mean duration of prescription was 33 weeks in the year of study, with a high of 38 weeks in group I-GERD. Almost one fourth of all patients (23%) had been using these drugs for more than 1 year. In more than half of all patients (53%), the medication was prescribed for 1 episode; in the other 47%, medication was prescribed for 2 or more episodes (ie, intermittent prescription). During the study period, 154 patients (17%) had used potential risk-bearing comedication for more than 6 weeks, including 48 with ulcer.
TABLE 1
CHARACTERISTICS AND PRESCRIPTIONS IN 922 PATIENTS (%) WITH LONG-TERM ACID SUPPRESSANT DRUG PRESCRIPTION IN 24 GENERAL PRACTICES IN THE REGION OF AMSTERDAM, THE NETHERLANDS
| DIAGNOSIS AFTER INVESTIGATION | |||||
|---|---|---|---|---|---|
| Characteristics | Total (N=922) | I-ULCER (n = 271) | I-GERD (n = 294) | I-FUNCTIONAL (n = 127) | NI: Stomach Complaints Without Investigation (n = 230) |
| Patients | |||||
| Female | 511 (55) | 41% | 57% | 72% | 62% |
| 15–44 years | 169 (18) | 11% | 17% | 25% | 25% |
| Mean age, years | 61 | 63 | 63 | 57 | 59 |
| Medication, No. | |||||
| Ranitidine | 442 (48) | 141 (52) | 115 (39) | 70 (55) | 116 (50) |
| Cimetidine | 236 (26) | 82 (30) | 48 (16) | 39 (31) | 67 (29) |
| Omeprazole | 241 (26) | 63 (23) | 130 (44) | 22 (17) | 26 (11) |
| Famotidine | 43 (5) | 16 (6) | 15 (5) | 4 (3) | 8 (3) |
| Lansoprazole | 13 (1) | 3 (1) | 5 (2) | 5 (4) | 0 (0) |
| Prescription Time | |||||
| 12–19 weeks | 231 (25) | 67 (25) | 44 (15) | 44 (35) | 76 (33) |
| 20–29 weeks | 184 (20) | 56 (21) | 46 (16) | 27 (21) | 55 (24) |
| 30–39 weeks | 148 (16) | 40 (15) | 50 (17) | 22 (17) | 36 (16) |
| 40–51 weeks | 143 (16) | 44 (16) | 60 (20) | 15 (12) | 24 (10) |
| >52 weeks | 216 (23) | 64 (24) | 94 (32) | 19 (15) | 39 (17) |
| Mean, weeks | 33 | 34 | 38 | 29 | 29 |
| No. of Episodes of Prescription | |||||
| 1 | 485 (53) | 144 (53) | 179 (61) | 51 (40) | 111 (48) |
| 2 | 271 (29) | 78 (29) | 81 (28) | 48 (38) | 64 (28) |
| > 2 | 166 (18) | 49 (18) | 34 (12) | 28 (22) | 55 (24) |
| Group I-ULCER includes all patients with a duodenal, -gastric, or nonspecified ulcer; group I-GERD includes all patients with symptomatic or erosive gastroesophageal reflux disease; group I-FUNCTIONAL includes patients with gastritis or with normal aspect on endoscopy or barium meal. | |||||
| *Total equals more than 100% because of different types of medication per patient; rarely prescribed medications are not mentioned. | |||||
Confirmation of working diagnosis
In 692 of the 922 (75%) patients a diagnostic test was performed to confirm the primary working diagnosis. In 519 (75%), a gastroscopy was performed and in 138 (20%) a barium meal radiograph was taken. In 35 patients, the specific form of investigation was unclear.
The specific diagnoses of the subgroups are shown in Table 2. In patients with ulcer, use of NSAIDs or prednisone was mentioned as the cause of the ulcer in 26 of 271 (9.6%) patients. Barrett’s esophagus was diagnosed in 29 of 342 (8.4%) patients in group I-GERD. In approximately 50% of the total number of patients, the investigation had been performed more than 5 years previously. Each patient had been treated accordingly during the subsequent years.