Access the full text.
Sign up today, get DeepDyve free for 14 days.
ContextAcute diarrhea with gas-related abdominal discomfort is a common, usually self-limited disorder with substantial social and economic impact.ObjectiveTo compare the efficacy and safety of a loperamide hydrochloride–simethicone combination product with those of loperamide alone, simethicone alone, and placebo in treating acute diarrhea with gas-related abdominal discomfort.DesignRandomized, placebo-controlled, double-blind trial of 48 hours' duration.SettingA primary care, ambulatory practice in Acapulco, Mexico.PatientsA total of 493 outpatient adults aged 18 to 63 years, with acute nonspecific diarrhea with at least moderately severe abdominal discomfort.InterventionsEach patient was randomly assigned to receive 2 chewable tablets containing loperamide hydrochloride, 2 mg, and simethicone, 125 mg (n=124); loperamide hydrochloride, 2 mg (n=123); simethicone, 125 mg (n=123); or placebo (n=123). This was followed by 1 tablet after each unformed stool, up to 4 tablets in any 24-hour period.Main Outcome MeasuresTime to last unformed stool and time to complete relief of gas-related abdominal discomfort were the protocol-specified primary outcomes. Secondary outcomes included time to complete relief of diarrhea, number of unformed stools, and patient-assessed variables at the end of the study (overall illness relief, diarrhea relief, and abdominal discomfort relief).ResultsPatients who received loperamide-simethicone had significantly (P<.001) shorter time to last unformed stool and faster relief of gas-related abdominal discomfort than patients who received loperamide, simethicone, or placebo alone. Loperamide-simethicone was significantly (P≤.01) more effective than the other 3 treatments for all end-of-study patient-assessed outcomes and all clinically important secondary outcomes. No significant differences in adverse events were found among treatment groups.ConclusionsThe loperamide-simethicone combination chewable product provides faster and more complete relief of acute nonspecific diarrhea and associated gas-related abdominal discomfort (gas pain, cramps, gas pressure, and bloating) than either of its components or placebo. The combination is well tolerated.ACUTE DIARRHEA is a common, usually self-limited disorder that occurs in a majority of adults at least once a year.Incidence rates of 0.53 to 0.55 per person-year have been reported for adults aged 20 to 39 years.There are multiple causes of acute diarrhea, including ingestion of toxins, food intolerance, psychological stress, and adverse reactions to medications. In most cases the cause is thought to be infections, including those caused by viruses, bacteria, or parasites.Many patients with acute diarrhea, regardless of cause, also experience gas, cramps, abdominal pain, bloating, distention, flatulence, nausea, and vomiting.It has been reported that 79% of patients with traveler's diarrhea have gas symptoms associated with their diarrheal illness.The social and economic impact of acute diarrhea can be substantial, resulting in days missed from work or school, interference with professional and personal plans and commitments, or the untimely disruption of travel or vacation plans. It has been estimated that 300 million people travel annually from one country to another.Of the estimated 16 million travelers from industrialized countries to developing countries,25% to 50% will develop diarrhea and 30% of those affected will be confined to bed.In contrast to diarrhea caused by invasive bacterial infections, pseudomembranous colitis, or idiopathic inflammatory bowel disease, acute nonspecific diarrheal illnesses are usually mild and self-limited, with symptoms resolving within 3 to 5 days. The majority of patients successfully obtain symptomatic relief of acute diarrhea with over-the-counter medications; however, until now there has been no single medication available that effectively relieved the symptoms of acute diarrhea and associated gas-related discomfort. Loperamide hydrochloride is an effective, safe antidiarrheal agent that was originally approved for prescription use in Europe in 1973 and in the United States in 1977; it has been available as an over-the-counter medication in the United States since 1988.Simethicone, a chemically inert compound that is not absorbed from the gastrointestinal tract, has been available in many countries since the 1960s for relief of gas-related abdominal discomfort. The current study was designed to compare the efficacy and safety of a loperamide and simethicone combination product with that of loperamide alone, simethicone alone, and placebo in treating acute nonspecific diarrhea with gas-related abdominal discomfort such as gas pain, cramps, gas pressure, and bloating.SUBJECTS AND METHODSPROTOCOLThis double-blind, placebo-controlled, outpatient study was conducted at one primary care, ambulatory practice in Acapulco, Mexico; this practice had several offices. Some study participants referred themselves for care because of general knowledge of the practice or because of advertisements for the study; others were referred by staff at hotels with satellite medical offices of the practice. The study was approved by an institutional review board, and all study participants gave written informed consent to participate.The inclusion criteria for participation in the study were as follows: (1) men or women 18 years or older; (2) acute diarrheal illness symptoms for less than 48 hours before study entry; (3) a minimum of 3 unformed stools passed within the 24 hours before study entry; (4) most recent bowel movement unformed; (5) gas-related abdominal discomfort (gas pain, cramps, gas pressure, or bloating) experienced during the hour before study entry; (6) maximum intensity of gas-related abdominal discomfort during the hour before study entry rated as either moderately severe or severe; and (7) if female, either postmenopausal or using an effective form of birth control for at least 3 months before study entry.Patients were excluded if they met any of the following conditions: (1) were suffering from a diarrheal illness severe enough to require hospital admission, outpatient parenteral hydration, or antibiotic therapy; (2) had an oral temperature greater than 38.6°C; (3) had a history or clinical evidence of gross blood or pus in the stool as part of the present illness; (4) exhibited signs or symptoms of orthostatic hypotension; (5) were unable to take medications and fluids by mouth; (6) had a history of chronic gastrointestinal tract disease, hepatic or renal insufficiency, or other major medical condition that might have been aggravated by untreated acute diarrhea; (7) had a history of hypersensitivity to loperamide or simethicone; (8) had taken any antibiotic or other drug known to substantially interfere with bacterial flora in the gut within 7 days before entry into the study; (9) had taken any antidiarrheal or promotility drug or antiflatulent within 12 hours before study entry; (10) had taken any analgesic within 6 hours before study entry; (11) were female and were pregnant, nursing, or experiencing perimenstrual abdominal or pelvic discomfort; or (12) had been previously enrolled in this study.Patients were not to take any other antidiarrheal or promotility drugs during the study period. In addition, they were not to take antacids, antiflatulents, antibiotics, or analgesics while participating in the study. Patients were advised to avoid consuming alcoholic beverages, carbonated beverages, nonpotable water, and foods and beverages containing milk or milk products during the study.Each patient was sequentially assigned to double-blind study medication (ie, numbered vials with attached opaque sealed envelopes containing the treatment assignment) according to the computer-generated randomization code allocated in blocks of 8. The assigned study medication was dispensed after eligibility had been confirmed and written informed consent had been obtained. The unopened treatment unblinding envelope was affixed to the patient's case report form.All 4 treatment groups received chewable tablets that were identical in appearance and taste. The treatment masking was kept intact and was not broken for any patient during the study. Neither the patients, the investigator, nor personnel directly involved in monitoring the study or reviewing the data knew the treatment assignments until the code was broken and the data were analyzed.Patients were randomly assigned to 1 of the following 4 treatment groups: loperamide hydrochloride, 2 mg, and simethicone, 125 mg, combination product; loperamide hydrochloride, 2 mg; simethicone, 125 mg; or placebo. Patients initially took 2 chewable tablets of study medication under the observation of study personnel. This was followed by administration of 1 tablet after each subsequent unformed stool, up to 4 tablets in any 24-hour period. Patients were dispensed a total of 8 tablets for the 48-hour study period. This dosing schedule accommodated the current recommended nonprescription dosing in the United States for both loperamide and simethicone.Patients recorded their symptoms, time and consistency of bowel movements, and time of administration of study medication during the 48-hour study period. For the first 8 hours of the study, patients rated the intensity of their gas-related abdominal discomfort hourly. Additional ratings were made 12, 24, 36, and 48 hours after the initial dose and each evening and morning during the study. At the end of the 48 hours or at the time of withdrawal from the study, patients recorded if and when they experienced complete relief from their diarrhea and their gas-related abdominal discomfort. Patients also evaluated and recorded end-of-study effectiveness of the study medication for overall illness relief, diarrhea relief, and abdominal discomfort relief by means of the following scale: 0, poor; 1, fair; 2, good; 3, very good; and 4, excellent. Patients returned to the study site for a second visit, 48 to 72 hours after the entry visit. Study personnel reviewed the completed diary with the patient to ensure that all required information had been recorded and to clarify any reported adverse events.Patients who withdrew from the study provided information on symptoms and end-of-study effectiveness at the time of withdrawal before self-medicating with nonstudy (rescue) medication.OUTCOME MEASURESThere were 2 primary outcome measures in this study. The time to last unformed stool, which objectively assessed the effect of the study drugs on reducing the duration of diarrheal symptoms during the 48-hour study period, was the primary efficacy outcome measure for assessing diarrheal symptom relief. The time to complete relief of gas-related abdominal discomfort, a subjective assessment of the treatment's effect on the duration of intestinal gas symptoms, was the primary efficacy outcome measure for evaluating gas symptom relief.In addition, there were several secondary outcome measures. The time to complete relief of diarrhea and the number of unformed stools were secondary outcome measures for assessing diarrheal symptom relief. Patient-assessed end-of-study effectiveness ratings for overall illness relief, diarrhea relief, and abdominal discomfort relief were also secondary outcome measures.STATISTICAL METHODSSample size estimates were obtained by means of parametric methods applied to pilot data, which indicated that a sample size of 480 (120 per treatment group) would be sufficient to detect significant differences between treatment group means of at least 7 hours at an α level of .05 (2-tailed) with a power of .80.Demographic and baseline characteristics were compared among treatment groups by means of a 1-way analysis of variance for continuous variables and by Fisher exact test for dichotomous variables. Fisher exact test was also used to compare adverse event rates.Survival analysis techniqueswere used to compare the treatment groups with respect to the time to relief for each of the primary outcome measures and for time to complete relief of diarrhea. Patients who did not complete the 48-hour study period were censored at the hour of withdrawal from the study. Both the Wilcoxon test and the log rank test were used to evaluate treatment group differences. These 2 tests produced comparable results for all evaluations.Analysis of variance was used to assess end-of-study effectiveness. Analysis of variance for repeated measures was used to assess the mean number of stools during each of the 4 successive 12-hour periods during the study.Analyses were conducted in 2 ways: (1) using an intent-to-treat approachand (2) including only the subset of patients who adhered to the protocol. Since these 2 approaches produced comparable results for all evaluations, only the results with the intent-to-treat approach are presented.All Pvalues reported are 2 sided.RESULTSPARTICIPANT FLOW AND FOLLOW-UPA total of 493 patients entered the study between September 10, 1993, and September 2, 1994. Two subjects became unavailable for follow-up before any safety and efficacy data were reported and were excluded from all efficacy and safety evaluations. Figure 1summarizes the randomization profile and the status of patients in the study; the number of patients who completed the study, the number who withdrew, and their primary reason for withdrawal are shown.Figure 1.Profile of the randomized controlled trial. Time to last unformed stool and time to complete relief of gas-related abdominal discomfort (gas pain, cramps, gas pressure, and bloating) were the protocol-specified primary outcomes.Table 1shows the demographic and baseline characteristics for all subjects enrolled in the study. The 4 treatment groups were comparable at baseline for demographic characteristics. Overall, 244 (49.5%) of patients enrolled in the study were men, the average age of the patients was 29.6 years, and all but 2 patients classified themselves as white (categories of race were white, black, and other; Hispanics were most likely counted as white). The average time from onset of diarrheal illness to the first dose of study treatment was 18.0 hours, and the average number of unformed stools in the 24 hours before treatment was 4.9. Significantly (P<.001) more patients assigned to the loperamide-simethicone group (16.9%) reported severe overall gas-related abdominal discomfort compared with the other 3 treatment groups at baseline (4.9%, 4.1%, and 1.6% for the loperamide, simethicone, and placebo groups, respectively). To address this imbalance, analyses of the 2 primary outcomes, time to last unformed stool and time to complete relief of gas-related abdominal discomfort, were repeated with exclusion of patients with severe overall gas-related abdominal discomfort at baseline. Results were similar to those reported below for all patients (data not shown).Table 1. Demographic and Baseline Characteristics of All Patients by Treatment GroupCharacteristicLoperamide Hydrochloride–Simethicone (n=124)Loperamide (n=123)Simethicone (n=123)Placebo (n=123)P*Sex, No. (%)Male55 (44)65 (53)58 (47)66 (54).40Female69 (56)58 (47)65 (53)57 (46)Race, No. (%)White123 (99)122 (99)123 (100)123 (100)Black1 (1)1 (1)00>.99Other0000Age, mean (SD), y28.9 (6.7)30.2 (7.8)29.4 (7.3)29.9 (6.6).52Treatment delay,† mean (SD), h18.9 (10.1)17.0 (10.3)17.7 (9.4)18.1 (8.8).46No. of unformed stools in previous 24 h, mean (SD)4.9 (1.3)4.8 (1.4)4.9 (1.3)5.0 (1.3).82Severe abdominal discomfort, No. (%)21 (17)6 (5)5 (4)2 (2)<.001*Statistical comparisons of categorical variables were performed with Fisher exact test; comparison of continuous means were performed with 1-way analysis of variance.†Treatment delay was defined as the time from onset of diarrheal illness to the first dose of study treatment.EFFICACY IN CONTROL OF DIARRHEAPatients assigned to the combination of loperamide and simethicone reported significantly (P<.001) faster relief for the primary diarrheal outcome measure of time to last unformed stool compared with patients assigned to either loperamide alone, simethicone alone, or placebo groups (Figure 2). Patients assigned to the combination of loperamide and simethicone had a reduction of 75% in the median time to last unformed stool compared with patients assigned to placebo (9.7 vs 39.0 hours), whereas patients assigned to loperamide alone or simethicone alone had reductions in the median time of 40% (23.4 vs 39.0 hours) and 17% (32.5 vs 39.0 hours), respectively, compared with placebo.Figure 2.Primary diarrheal outcome: time to last unformed stool by treatment group. Wilcoxon test, P<.001 for loperamide hydrochloride–simethicone vs each of the other treatment groups.Similar results were observed for the secondary diarrheal outcome measure, number of unformed stools (Table 2). Overall, patients in the loperamide-simethicone group had a significantly (P<.001) lower mean number of unformed stools during each of 4 periods (>0-12 hours, >12-24 hours, >24-36 hours, and >36-48 hours) compared with patients in the simethicone or placebo groups, and a significantly (P=.01) lower mean number of unformed stools during the first period (>0-12 hours) compared with patients in the loperamide group. The mean number of unformed stools was somewhat lower for patients treated with the combination of loperamide and simethicone compared with loperamide alone during the later periods (>12-24 hours, >24-36 hours, and >36-48 hours); however, these differences were not statistically significant. Given the self-limited nature of acute nonspecific diarrhea, the initial period (>0-12 hours) is of greatest clinical importance.Table 2. Secondary Diarrheal Efficacy Outcome Measure: Mean Number of Unformed Stools by Treatment GroupInterval Since Treatment, hMean (SEM)*Loperamide Hydrocholoride– Simethicone (n=124)Loperamide (n=123)Simethicone (n=123)Placebo (n=123)>0-121.04 (0.08)1.40 (0.08)†2.36 (0.08)‡2.29 (0.08)‡>12-240.36 (0.08)0.60 (0.08)1.25 (0.08)‡1.31 (0.08)‡>24-360.47 (0.08)0.69 (0.08)1.94 (0.08)‡2.03 (0.08)‡>36-480.12 (0.08)0.25 (0.08)0.76 (0.08)‡1.14 (0.08)‡*Least squares mean.†P=.01 compared with loperamide hydrochloride–simethicone group.‡P<.001 compared with loperamide hydrochloride–simethicone group.As shown in Figure 3, the time to complete relief of diarrhea was significantly (P<.001) shorter for the loperamide-simethicone group than for the loperamide group, the simethicone group, or the placebo group.Figure 3.Secondary diarrheal outcome: time to complete relief of diarrhea by treatment group. Wilcoxon test, P<.001 for loperamide hydrochloride–simethicone vs each of the other treatment groups.EFFICACY IN RELIEF OF GAS-RELATED ABDOMINAL DISCOMFORTPatients assigned to the combination of loperamide and simethicone reported significantly (P<.001) faster relief for the primary outcome measure for gas-related symptom relief efficacy, time to complete relief of gas-related abdominal discomfort (gas pain, cramps, gas pressure, and bloating), than either loperamide alone, simethicone alone, or placebo (Figure 4). Patients assigned to the combination of loperamide and simethicone had a reduction of 75% in median time to complete relief of gas-related abdominal discomfort compared with patients assigned to placebo (12.0 vs 48.0 hours), whereas patients assigned to loperamide alone or simethicone alone had reductions in the median time of 13% (42.0 vs 48.0 hours) and 56% (21.1 vs 48.0 hours), respectively, compared with placebo.Figure 4.Primary gas-related symptom relief outcome: time to complete relief of abdominal discomfort by treatment group. Wilcoxon test, P<.001 for loperamide hydrochloride–simethicone vs each of the other treatment groups.END-OF-STUDY EFFECTIVENESSPatients randomized to the combination of loperamide and simethicone reported a significantly (P<.001) higher mean end-of-study rating of overall illness relief (3.0), diarrhea relief (2.9), and abdominal discomfort relief (3.0) compared with patients randomized to loperamide, simethicone, or placebo (Figure 5). Patients randomized to loperamide reported a higher mean end-of-study rating of diarrheal relief (2.3) than patients assigned to simethicone (1.1). However, patients randomized to simethicone reported a higher mean end-of-study rating of abdominal discomfort relief (1.9) than patients assigned to loperamide (1.5).Figure 5.Secondary efficacy outcomes: mean patient-assessed end-of-study effectiveness evaluation (0, poor; 4, excellent) by treatment group. Based on analysis of variance for each of these 3 outcomes, P<.001 for loperamide hydrochloride–simethicone vs each of the other treatment groups.WITHDRAWAL OF STUDY MEDICATION AND SAFETYForty patients were withdrawn prematurely from study medication. Two of these patients became unavailable for follow-up before any follow-up data were collected. In addition, 7 patients (4 assigned to loperamide-simethicone and 3 assigned to loperamide) withdrew from treatment because diarrhea resolved within the first day of treatment. In an additional 22 patients (10 assigned to simethicone, 11 assigned to placebo, and 1 assigned to loperamide), the need for rescue medication led to early withdrawal of study medication. Other reasons for withdrawal included taking medication for concomitant illness, adverse event, and patient request.The number of patients reporting adverse events was comparable among the 4 treatment groups (6 patients assigned to placebo, 8 patients assigned to loperamide-simethicone, 5 patients assigned to loperamide, and 2 patients assigned to simethicone; P=.27, Pvalue from Fisher exact test comparing adverse event rates over all treatments). No serious adverse events were reported. Adverse events reported by more than 1 patient included the following: headache (4 patients: 1 loperamide-simethicone, 3 placebo), nausea (5 patients: 4 loperamide-simethicone, 1 loperamide), cough (3 patients: 2 loperamide-simethicone, 1 placebo), and rash (2 patients in the loperamide group).Of the 3 patients who withdrew from the study because of adverse events, 1 in the simethicone group had lumbar pain, and in the placebo group 1 had rheumatic pain and 1 had cough and pharyngitis; these adverse events were considered to be unrelated to the study medication.COMMENTA single, well-tolerated, self-administered product that effectively relieves the symptoms of diarrhea and the gas-related abdominal discomfort that commonly occurs with it, would be useful and welcome for patients suffering from this condition. The active ingredients in the combination product studied, loperamide and simethicone, have been available over the counter individually for a number of years.It should be noted that antimotility agents are not recommended for patients with acute diarrhea with high fever or blood or mucus in the stool; these patients likely require antimicrobial therapy.Treatment with loperamide-simethicone is indicated for the symptoms of acute diarrhea with gas-related abdominal discomfort when fever is absent or low grade and dysentery is not present. Oral rehydration and changes in diet are recommended for all patients with diarrhea.In this study, the combination of loperamide and simethicone was shown to provide faster relief of acute nonspecific diarrhea and the gas-related discomfort (gas pain, cramping, gas pressure, and bloating) associated with it than either ingredient alone. This was an unexpected finding. The combination product proved to be more effective than either ingredient alone or placebo for both primary outcome measures, all clinically important secondary outcome measures, and all 3 patient-assessed efficacy outcome measures at the end of the study. Moreover, study withdrawals associated with the need for rescue medication were almost exclusively in the placebo and simethicone treatment groups. Patient satisfaction, the outcome required ultimately of all over-the-counter medications, as measured by end-of-study patient-assessed outcome measures, was significantly in favor of the combination product over either ingredient alone or placebo. The combination product offers faster control of diarrhea not only when compared with placebo and simethicone but also when compared with loperamide alone. The combination also offers faster relief of associated abdominal discomfort not only when compared with placebo and loperamide but also when compared with simethicone. Interestingly, these data demonstrate the efficacy of simethicone alone compared with placebo in the treatment of abdominal discomfort associated with diarrhea.The results obtained from this trial clearly indicate that use of the loperamide-simethicone combination product leads to an enhancement in both diarrheal symptom relief and gas-related symptom relief beyond that seen with each individual component. The exact mechanism for this observed augmented response is not known. Loperamide's ability to decrease gut motility, increase fluid reabsorption, and decrease intestinal secretion could conceivably enhance simethicone's defoaming activity by increasing the contact time on a reduced volume of intestinal fluid. It has been suggested that optimization of loperamide's antidiarrheal activity is obtained when low luminal concentrations of loperamide are distributed over the greatest length of the small intestine.Using human pharmacokinetic data, a reduction in the rate of loperamide absorption (peak plasma concentration and time to maximal plasma concentration) has been demonstrated for the loperamide-simethicone combination product when compared with loperamide capsules (unpublished data on file, Thomas L. Hunt, MD, PhD, 1995). These changes in loperamide absorption suggest that simethicone may interact with loperamide, leading to a slow, steady release of loperamide. This could potentially result in the availability of low concentrations of loperamide for a longer period to a larger portion of small intestinal mucosal receptors and thus produce an increase in antidiarrheal activity.In summary, these results indicate that the loperamide-simethicone combination chewable product provides faster and more complete relief of both acute nonspecific diarrhea and the gas-related abdominal discomfort (gas pain, cramps, gas pressure, and bloating) associated with it than either of its components or placebo.HJBrownleeFamily practitioner's guide to patient self-treatment of acute diarrhea.Am J Med.1990;88(suppl 6A):27S-29S.ASMontoJSKoopmanThe Tecumseh Study, XI: occurrence of acute enteric illness in the community.Am J Epidemiol.1980;112:323-333.SIParkRAGiannellaApproach to the adult patient with acute diarrhea.Gastroenterol Clin North Am.1993;22:483-497.RLGuerrantDABobakBacterial and protozoal gastroenteritis.N Engl J Med.1991;325:327-340.NRBlacklowHBGreenbergViral gastroenteritis.N Engl J Med.1991;325:252-264.PHKatelarisMJGFarthingTraveler's diarrhea: clinical presentation and prognosis.Chemotherapy.1995;41(suppl 1):40-47.SLGorbachInfectious diarrhea and bacterial food poisoning.In: Sleisinger MH, Fordtran JS, eds. Gastrointestinal Disease.5th ed. Philadelphia, Pa: WB Saunders Co; 1993:1128-1173.RSteffenEpidemiologic studies of travelers' diarrhea, severe gastrointestinal infections, and cholera.Rev Infect Dis.1986;8(suppl 2):S122-S130.Consensus Development Conference PanelConsensus Development Conference Statement.Rev Infect Dis.1986;8(suppl 2):S227-S233.SLGorbachTravelers' diarrhea.N Engl J Med.1982;307:881-883.GEDukesOver-the-counter antidiarrheal medications used for the self-treatment of acute nonspecific diarrhea.Am J Med.1990;88(suppl 6A):24S-26S.DCollettModelling Survival Data in Medical Research.London, England: Chapman & Hall; 1994.SJPocockClinical Trials: A Practical Approach.Chichester, England: John Wiley & Sons; 1983.HLDuPontthe Practice Parameters Committee of the American College of GastroenterologyGuidelines on acute infectious diarrhea in adults.Am J Gastroenterol.1997;92:1962-1975.FAwoutersAMegensMVerlindenJSchuurkesCNiemegeersPJanssenLoperamide: survey of studies on mechanism of its antidiarrheal activity.Dig Dis Sci.1993;38:977-995.Accepted for publication June 4, 1998.This study was conducted, analyzed, and supported by McNeil Consumer Healthcare, Fort Washington, Pa.We are grateful for the efforts of Esteban Ortiz, MD, and his staff, whose patients participated in this study.Reprints: Michael A. Kaplan, MD, Medical Department, McNeil Consumer Healthcare, 7050 Camp Hill Rd, Fort Washington, PA 19034 (e-mail: firstname.lastname@example.org).
Archives of Family Medicine – American Medical Association
Published: May 1, 1999
Access the full text.
Sign up today, get DeepDyve free for 14 days.