Aspirin and nonaspirin nonsteroidal antiinflammatory drugs (NSAIDs) are among the most commonly used medications. Prescription use of NSAIDs in the United States appears to be stabilizing, but nonprescription (over-the-counter, OTC) use is growing. Sales of ibuprofen – Advil, Motrin IB, Nuprin – have more than tripled since the analgesic was approved for OTC sales nearly a decade ago. Overall sales of OTC NSAIDs will get another boost now that naproxen (Aleve) has received OTC marketing approval.
The increase in over-the-counter (OTC) sales comes as no surprise. Buying an analgesic off the shelf is considerably easier and more convenient – and far less expensive – than seeing a doctor and going through the process of getting a prescribed drug. OTC medications are widely promoted and readily available. Unfortunately, the average American does not realize that a drug purchased OTC may be associated with the same adverse effects as the same drug purchased by prescription. In the case of OTC NSAIDs, adverse reactions may include kidney damage, hypertension, and gastrointestinal (GI) symptoms ranging from mild dyspepsia to serious or even fatal GI bleeding.
NSAID-induced GI Disease
GI disease associated with NSAIDs has been reported as the most common serious adverse drug effect in the United States. Nonsteroidal antiinflammatory drugs (NSAIDs) can exacerbate underlying disease or cause new lesions. From 38% to 50% of patients taking NSAIDs report dyspepsia, and gastric ulcers may develop in up to 28% of patients taking NSAIDs regularly. Studies have shown that a person exposed to NSAIDs has three to four times the risk of non-users for upper GI bleeding, perforation, or both. Increased risk of a serious adverse reaction has been associated with age, female sex, and rheumatoid arthritis, while independent risk factors associated with upper GI bleeding include male sex, history of peptic ulcer (with or without complications), and the use of alcohol, cigarettes, anticoagulants, and corticosteroids.
Lands et al. found that 80% of patients with either upper or lower GI bleeding, and 78% of patients with upper GI bleeding only, had recently consumed NSAIDs (verified by measuring platelet cyclooxygenase activity and serum salicylate levels, as well as taking patient history). Klein et al. also found a higher frequency of GI bleeding in NSAID users. They examined discharge data from hospitalized patients who had suffered GI hemorrhage (both NSAID users and nonusers). Patients using nonsteroidal antiinflammatory drugs (NSAIDs) spent more time in intensive care than nonusers (median 1 day versus 0 days), and daily users had a higher transfusion requirement than nonusers (4 units versus 1 unit), both of which suggest that NSAID use has a substantial impact on health-care resource allocation.
OTC NSAIDs and GI Hemorrhage
A two-year study of hospitalized patients admitted with upper GI bleeding has demonstrated that the self-administration of OTC NSAIDs can substantially increase the risk of bleeding. Wilcox et al. evaluated consecutive patients admitted with upper GI hemorrhage to a large inner-city hospital in Atlanta, Georgia. The use of any OTC or prescription nonsteroidal antiinflammatory drug (NSAID) during the week before admission was assessed prospectively (from patient and family interview and pharmacy records). Inclusion criteria included age more than 18 years; presence of hematemesis and a subnormal hematocrit (or a decrease in hematocrit of more than 5 points from baseline); and documented lesion (disclosed by endoscopy, barium upper GI series, or at autopsy). Patients were excluded if the GI bleeding first occurred during hospitalization.
During the 2-year period, 421 patients were evaluated. Mean age was 50 (range 18-89); 352 patients were black, 63 white, 4 Hispanic, and 2 Asian; 276 patients were male, and 145 were female. Peptic ulcer was the most common cause of bleeding, identified in more than half the patients. Acute gastric mucosal lesions – including nonspecific gastritis, portal hypertension, and gastropathy (alcohol- or aspirin- induced) – were relatively infrequent. More women than men and more subjects over age 60 than under 60 were taking prescription NSAIDs; those younger than 60 were more likely to use OTCs, and those over 60 were more likely to use prescription NSAIDs. Drug use was not related to race.
Aspirin was taken during the week before admission by 41% of patients (and OTC aspirin by a total of 35%). OTC nonaspirin-NSAIDs were taken by 9% of patients and prescription NSAIDs by 14%. The investigators described the use of OTC aspirin and nonaspirin NSAIDs as “striking” in patients with upper GI bleeding (all causes). These products were used most frequently by patients with ulcer-related bleeding (66%), esophagitis (62%), and Mallory-Weiss tears (laceration of the gastric cardia) (56%) but were also used commonly by patients with bleeding unrelated to ulcers. The investigators concluded, “We believe, as do others, that short-term NSAID use (less than 1 week) may be one of the most important precipitating factors for ulcer related hemorrhage.”
Epidemiology of NSAID-GI Bleeding
A large retrospective study of NSAID use and upper GI (gastrointestinal) bleeding has illuminated the use and misuse of nonsteroidal antiinflammatory drugs (NSAIDs). Rodriguez et al. used a database of 4 million patients (the UK General Practitioners’ Computerized Records) to identify 1467 patients with upper GI bleeding over a 3-year period, and 10,000 matched controls. The site of bleeding was gastric in 483 patients and duodenal in 787; 40 patients had multiple sites of bleeding, 147 had peptic ulcer only, and 261 had perforation; 64 patients died. One major indication for NSAID therapy was ill-defined back pain (13% for cases and 10% for controls). Although patients used any of 17 different NSAIDs, sufficient data for individual analyses were obtained for ibu- profen, naproxen, diclofenac, ketoprofen, indomethacin, piroxicam and azapropanone (the latter not available in the United States). Aspirin use was not specifically addressed, because use was seen as too widespread and underreported.
The investigators found that the mean relative risk for upper GI bleeding associated with current NSAID use was 4.7 (7.0 with high doses and 2.6 with low doses). Current users were patients who most likely used nonsteroidal antiinflammatory drugs (NSAIDs) within the previous month (last NSAID prescription ordered during the month before the index date, or duration of therapy including the index date). NSAIDs associated with highest risk of GI bleeding were azapropanone (relative risk 23.4) and piroxicam (relative risk 18). Ibuprofen at low doses had the smallest risk (2.1), although risk was substantially increased with higher doses (6.5 for 1500 mg/day or more). Both short- and long-duration exposure increased the risk of upper GI bleeding, long duration only slightly more than short duration. Patients who recently switched from one NSAID to another or used more than one NSAID simultaneously had more than twice the risk of patients exposed to only one NSAID. In cases where current aspirin use was recorded on the computer, adjustment did not greatly alter risks, which differs from the other studies.
The biggest risk factor for GI bleeding in NSAID users was a history of peptic ulceration (relative risk 17.2). Age also played a role: people under 60 exposed to NSAIDs had a relative risk of 2.8; people over 60 had a relative risk of 13.2. NSAID use was slightly greater in women than in men, yet male sex was associated with a greater risk. The investigators described the patient at greatest risk of presenting with an episode of upper GI (gastrointestinal) bleeding as “a male smoker of advanced age who has a history of peptic ulcer, and is a user of oral corticosteroids, anticoagulants, and NSAIDs.”
Managing NSAID-induced Dyspepsia
How should patients with NSAID-induced dyspepsia be managed? A prospective study by questionnaire (sent to 300 general practitioners, 261 of whom responded) found that management strategies include NSAID discontinuation (87%), switching to another NSAID (12%), and referral for endosco-py (14%) or barium meal (2%). The drugs preferred for management of dyspepsia included histamine2-receptor blockers (41%), antacids (about 25%), and misoprostol (25%).
The authors of the report offered the following guidelines for patients with persistent NSAID- related dyspepsia when simple analgesics are inadequate: Endoscopy is advisable when symptoms persist for more than 3 weeks in patients who need long-term NSAID therapy; when dyspeptic symptoms or signs suggest organic disease; and when the patient is receiving both steroids and NSAIDs. Misoprostol prophylaxis for NSAID-related dyspepsia is recommended at the onset of NSAID therapy for patients with a history of peptic ulcer disease and may also be justified if NSAID therapy is short term, or if endoscopy is intolerable, impractical, or unavailable.
NSAIDs and hypertension in the elderly.
Do NSAIDs alter blood pressure? In elderly patients, the answer to this question is a qualified “maybe.” Johnson et al. found NSAID use an independent risk factor for hypertension in nearly 3000 Australian subjects aged 60 years or more who answered a questionnaire and underwent blood pressure mea- surements. Frequency of NSAID usage was determined for all study participants – 1237 men and 1568 women stratified by age, sex, blood pressure group, and history of arthritis.
Nonsteroidal antiinflammatory drug (NSAID) usage was 26% overall (females 28% and males 23%). Usage increased with age and was higher in females than males in every group studied. Among patients with a history of “arthritis,” 45% were using NSAIDS; 12% were taking both NSAIDS and antihypertensive agents. NSAID usage significantly predicted the presence of hypertension, with an attributable risk of 29%. It is difficult to tell whether NSAIDs raise blood pressure or antagonize the effects of antihypertensive agents, said the investigators, however, “In this elderly population… 29% of the hypertension occurring amongst those taking NSAIDs could be attributed to the NSAID, underscoring the potential contribution of NSAID usage to the prevalence of hypertension in the elderly.”