Implementation of a nursing home urinary incontinence management program with and without tolterodine.

OBJECTIVE: To evaluate the implementation of a nursing home urinary incontinence management program. DESIGN: A prospective field trial of the program incorporating practice guidelines and principles of continuous quality improvement. SETTING: Five nursing homes in New York, Virginia, and Georgia PARTICIPANTS: One hundred fifty-one residents identified as being incontinent of urine and who met inclusion criteria for ongoing participation in the program. INTERVENTION: Key multidisciplinary staff from the five nursing homes were trained in the program and assumed responsibility for implementing it in their facilities. The program consisted of a clinical assessment, toileting protocols, and the addition of the antimuscarinic drug tolterodine in selected residents who did not respond well to toileting alone. Data on dryness rates during the 60-day toileting protocols, collected by nursing home staff, were analyzed on a weekly basis by an overall project coordinator who sent data back to the nursing homes in an easy-to-read graphical format. MEASURES: (1) The dryness rate, defined as the number of times the resident was dry divided by the number of times the resident was checked (every 2 hours from 7 a.m. to 7 p.m.); and (2) adverse events (eg, dry mouth, increased confusion, need for dosage reduction). RESULTS: Of 645 residents in the 5 nursing homes, 377 (58%) were identified as incontinent of urine, of whom 151 (40%) were placed on an ongoing toileting program. Of these 151 residents, 48 (32%) were prescribed tolterodine, and 117 (78%) completed the 60-day trial. The initial dryness rate was 57%, and for the group as a whole remained essentially unchanged (increase in dryness 1%, P = 0.50). Among 50 clinically stable residents on a toileting program alone, the increase in the dryness rate was 16% (P = 0.001), and for 31 clinically stable residents prescribed tolterodine, the increase in the dryness rate was 29% (P = 0.012). Two residents had their dosage of tolterodine reduced because of dry mouth and nausea,one resident was taken off the drug because of increased pain in the back and legs and increased confusion. CONCLUSIONS: Overall, this program resulted in significant increases in dryness rates for clinically stable incontinent nursing home residents. These residents represented 22% of the total number of residents identified as incontinent in the five participating nursing homes. Tolterodine was prescribed for approximately one-third of incontinent residents as a supplement to a toileting program, and was well tolerated. Nursing homes should be encouraged to implement similar urinary incontinence programs, target toileting protocols to the most responsive residents, and maintain the program using principles of continuous quality improvement.

Applying epidemiology-based outcomes research to clinical decision-making. A hypothetical model of biotechnology therapy in gram-negative sepsis.

OBJECTIVE: Sepsis occurs in a heterogeneous population. A prospective nationwide surveillance study found that populations stratified by infection type had significant differences in the incidence of sepsis syndrome, rate of complications and mortality. The objective of this study was to explore whether successful identification of population-specific risk factors for disease-associated morbidity and mortality may allow for more accurate assessment of the cost effectiveness of treatment strategies. DESIGN: A decision analytic model was developed using outcomes data on incidence and resolution of major complications in sepsis syndrome. Healthcare resource utilisation data were based on length of hospital stay, intensive care unit stay versus hospital ward stay, and cost of treating sepsis-related complications. SETTING: This modelling study, conducted from the perspective of the healthcare institution, used actual outcomes data on 2 infection-specific patient populations. PATIENTS AND PARTICIPANTS: The 2 populations studied were patients with nosocomial respiratory tract infection or community-acquired urinary tract infection who subsequently developed sepsis syndrome. INTERVENTIONS: Treatment options modelled were standard therapy plus biotechnology therapy versus standard therapy alone in the treatment of gram-negative sepsis complications. MAIN OUTCOME MEASURES AND RESULTS: The incremental cost-effectiveness ratios differed between the 2 study populations, due to differences in the incidence and rate of resolution of major sepsis-associated complications. The use of biotechnology therapy is always more cost effective in the respiratory tract infection population. CONCLUSIONS: Cost-effectiveness results for a therapy may change when the epidemiology of the disease state is known and incorporated into the decision analytic model. An infection-specific approach is important in the treatment of sepsis.

Symptomatic cardiotoxicity associated with 5-fluorouracil.

A prospective cohort study was conducted in 35 hospitals with oncology units to determine the incidence of symptomatic cardiotoxicity in patients receiving continuous infusions of 5-fluorouracil (5-FU), and to identify risk factors that could contribute to the development of 5-FU-associated cardiotoxicity. A sample of 483 patients [197 (41%) women, overall average age +/- SD 60.9 +/- 11.9 yrs] were followed for one cycle of 5-FU infusion. Thirty-eight (7.9%) patients had abrupt termination of the infusion. There were 9 (1.9%) cases of suspected or documented cardiotoxic events. Cardiotoxicity occurred in 7 (3.35%) of 209 patients receiving their first course of 5-FU and in 2 (0.73%) other patients (p=0.044). Based on univariate analysis, the following patient groups were at elevated risk of cardiotoxicity: those with preexisting cardiac disease (RR=6.83, p=0.0023); patients receiving calcium channel blockers (RR=4.75, p=0.014); those receiving nitrates (RR=9.18, p=0.007); and patients receiving concomitant etoposide (RR=10.32, p=0.022). Patients with underlying cardiac disease require close monitoring while receiving continuous infusions of 5-FU. They should be observed for signs and symptoms of cardiotoxicity, and vital signs should be measured frequently. Continued reporting of 5-FU-associated cardiotoxicity is necessary to identify other patients at risk.

Pretreatment regimens for adverse events related to infusion of amphotericin B.

Infusion-related adverse events (IRAEs) such as nausea, vomiting, fever, chills, and thrombophlebitis that are associated with amphotericin B therapy often lead clinicians to prescribe a number of adjunctive pretreatment medications in an attempt to reduce the incidence and severity of these events. The purpose of this study was to determine the incidence of IRAEs during the first week of systemic amphotericin B therapy and to identify pretreatment regimens that are effective in preventing these IRAEs. Three hundred ninety-seven adult inpatients receiving amphotericin B therapy were prospectively monitored, and data regarding IRAEs and pretreatment regimens were collected. Of these patients, 282 (71%) developed at least one IRAE during the first 7 days of therapy. The IRAEs most commonly reported were fever (51% of patients) and chills (28%), followed by nausea (18%), headache (9%), and thrombophlebitis (5%). The most common regimens included diphenhydramine, a corticosteroid, acetaminophen, and heparin, administered alone or in combination with these or other drugs. Overall, common pretreatment regimens were similar in efficacy to no pretreatment in the prevention of IRAEs. Thus empirical premedication for IRAEs associated with amphotericin B cannot be routinely advocated; instead, patients should be treated when symptoms first arise and then premedicated for subsequent amphotericin B infusions.

Adverse drug reaction reporting in a multicenter surveillance study.

OBJECTIVE: To evaluate the performance of a multicenter, prospective surveillance program in identifying adverse events, and to seek explanations for misclassification bias. DESIGN: The design was a prospective observational study of patients with documented or suspected bacterial pneumonia. SETTING: Data were collected in 74 acute care hospitals across the US. PATIENTS: This evaluation was based on a consecutive sample of 1822 adult patients (> 18 years of age) with documented or suspected bacterial pneumonia who were being treated with a cephalosporin, a penicillin, or an aminoglycoside over a 3-month period. Patients were followed for the duration of antibiotic therapy and were excluded if antibiotic therapy was < 3 days or if the pneumonia was judged to be nonbacterial. INTERVENTIONS: Clinical pharmacists documented patient demographics, concurrent illnesses and medications, antibiotic administration, relevant laboratory data, and the occurrence of nephrotoxicity and neutropenia. MAIN OUTCOME MEASURES: Validity of investigators' identification of neutropenia and nephrotoxicity as compared with objective laboratory data was assessed by using sensitivity, specificity, and positive and negative predictive value measures. RESULTS: Among the 1502 patients with sufficient data to evaluate neutropenia, there was agreement in 1270 patients (84.6%); likewise, among 1291 patients with sufficient data to evaluate nephrotoxicity there was agreement in 1186 patients (91.9%). Sensitivity of the researchers' assessments was 50.9% and 71.0% for neutropenia and nephrotoxicity, respectively. The negative predictive value was > 95% for both events. CONCLUSIONS: Overall, this evaluation demonstrated that the Drug Surveillance Network can successfully identify targeted adverse events. Moreover, this study highlights the importance of validation for all types of outcomes-oriented research studies.

Sepsis syndrome and associated sequelae in patients at high risk for gram-negative sepsis.

We conducted a prospective surveillance study of 80 hospitals across the United States to determine the incidence of sepsis syndrome and its associated sequelae in hospitalized patients over age 18 years who were administered antibiotics for suspected or documented gram-negative infection. A sample of 1754 hospitalized patients were followed from onset of antimicrobial therapy to discharge or death. Mortality rates (MR) varied depending on the suspected source of sepsis syndrome. For patients in whom the syndrome was associated with community-acquired urinary tract infections, mortality was 20% (relative risk [RR] = 0.51, p < 0.05), for those with trauma 20.6% (RR = 0.51, p < 0.05), and patients with nosocomial respiratory tract infections 57.1% (RR = 1.66, p < 0.05). More than two complications occurred in 65.2% of patients under age 60 years (MR 31%), 40.8% of those age 60-80 (MR 42%), and 35.6% of patients older than 80 years (MR 33.3%, p > 0.05). Various patient populations had significant differences in both the incidence of the syndrome and its complications, and consequent mortality. Perhaps morbidity as well as mortality should be used as outcomes when testing the efficacy of innovative therapies for sepsis.

Analysis of potential risk factors associated with the development of pancreatitis in phase I patients with AIDS or AIDS-related complex receiving didanosine.

Phase I dose-escalating trials of didanosine revealed dose-limiting toxicities, including pancreatitis, and established a total daily dose of 12.5 mg/kg/day as the maximum tolerated dose. Clinical and pharmacokinetic data of 61 patients from two trials were analyzed to further evaluate the risk of pancreatitis: 1 (6.3%) of 16 patients who received < 500 mg/day didanosine, 2 (13.3%) of 15 who received 500-750 mg/day, and 15 (50%) of 30 who received > 750 mg/day developed pancreatitis (P < .001). A relationship between risk of pancreatitis and steady-state plasma concentrations of didanosine and age was also observed, suggesting that knowledge of didanosine pharmacokinetics provided additional information regarding risk of toxicity. Further confirmation of these findings will be necessary to determine if the risk factors for pancreatitis remain the same at lower doses currently used.

Risk factors for acute renal insufficiency in patients with suspected or documented bacterial pneumonia.

OBJECTIVE: To describe the incidence of acute renal insufficiency and identify potential risk factors associated with this adverse medical event. DESIGN: A cohort analytic study of patients with documented or suspected bacterial pneumonia. SETTING: Nationwide survey of 74 acute care hospitals across the US. INCLUSION AND EXCLUSION CRITERIA: A total of 1822 adult patients with documented or suspected bacterial pneumonia who were receiving a cephalosporin, penicillin, or an aminoglycoside were enrolled. Patients were excluded if the duration of antimicrobial therapy was < 3 days or if the pneumonia was judged to be nonbacterial. DATA COLLECTION: Clinical pharmacists completed standardized data collection forms on all patients enrolled in the study. Information regarding patient demographics, concurrent illnesses and medications, antibiotic administration, representative laboratory data, and the occurrence of any adverse clinical event was specifically captured. Information regarding the development of acute renal insufficiency was targeted as an event to be captured. MAIN OUTCOME MEASURES: Univariate and multivariate analyses were performed to identify significant risk factors for acute renal insufficiency. A subset analysis was similarly performed to identify risk factors associated with aminoglycoside-related acute renal insufficiency. RESULTS: Of the patients enrolled in this study, 8.2 percent developed acute renal insufficiency. Risk factors for acute renal insufficiency included renal disease, aminoglycoside therapy, nosocomial pneumonia, elevated estimated creatinine clearance prior to study entry, cardiac arrest/shock, congestive heart failure, total duration of antibiotics > 7 days, clindamycin therapy, liver disease, and first-generation cephalosporin usage. Risk factors for aminoglycoside-related acute renal insufficiency identified via multiple logistic regression included amphotericin B, congestive heart failure, aminoglycoside trough concentration > 1.5 mg/L, and clindamycin therapy. CONCLUSIONS: The risk factors identified for acute renal insufficiency suggest that severity of illness strongly influences the development of renal insufficiency. Theoretically, the results of this study could serve as a framework for developing risk prevention programs within individual hospitals. Specific risk factors could be identified for a patient population and risk factors that could be modified could then be targeted for intervention. This type of information can also assist clinicians in predicting the probability of the adverse event for a particular patient and subsequently minimizing this risk by initiating intense monitoring or modifying the drug regimen.

Use of antifungal therapy in hospitalized patients. I. Results prior to the marketing of fluconazole.

OBJECTIVE: To evaluate the use of antifungal agents in hospitalized patients prior to marketing of fluconazole and to assess characteristics associated with their use. DESIGN: A cohort of hospitalized patients receiving topical or systemic antifungal therapy was monitored concurrently. SETTING: Sixty-nine hospitals ranging in size from 100 to more than 500 beds, 70.1 percent affiliated with medical schools. PATIENTS: Participating clinical pharmacists each identified 15 consecutive patients receiving systemic antifungal therapy and 5 consecutive patients receiving topical antifungal therapy at their institutions. Data collection began October 1989 and ended March 1990. INTERVENTION: All data collected were observational in nature, and no patient intervention was required. MEASURES: Characteristics of patients receiving antifungal therapy were compared using t-tests and chi-square tests. Utilization and patterns of use of antifungal therapy were reported. RESULTS: The most common risk factors necessitating antifungal therapy, in descending order, were: administration of broad-spectrum antibiotics and/or presence of invasive catheters, carcinoma, AIDS, leukemia or lymphoma, diabetes mellitus, solid organ or bone marrow transplantation, and chronic obstructive pulmonary disease. Five hundred seventeen patients received systemic therapy and 464 (89.7 percent) received a single systemic agent. Of these, 242 (52.2 percent) received amphotericin B, 215 (46.3 percent) received ketoconazole, 6 (1.3 percent) received flucytosine, and 1 (0.2 percent) received intravenous miconazole. Fifty-three patients received two systemic agents either concurrently or consecutively. Ketoconazole was most often used for presumed or documented oral, urogenital, or esophageal infections and amphotericin B was the preferred agent for disseminated infections and fungemia (p < 0.001). Almost half of the patients receiving amphotericin B or ketoconazole (48.3 percent) received these drugs as empiric therapy. Documented infections were more likely to be treated with amphotericin B (54.8 percent) than with ketoconazole (27.4 percent) (p < 0.001). The predominant fungal isolates were Candida albicans, Candida spp., and unspecified yeasts. Amphotericin B toxicity led to discontinuation of drug therapy in only 5.1 percent of cases. Two hundred sixty-nine patients (34.2 percent) received topical antifungal therapy only. Nystatin oral suspension was prescribed to 65.3 percent of the patients, clotrimazole troches to 23.0 percent, amphotericin B irrigation to 10.9 percent, and nystatin tablets to 0.8 percent. CONCLUSIONS: The utilization patterns of antifungal agents in this survey follow established therapeutic guidelines. Prior to the introduction of fluconazole, amphotericin B was the agent of choice for documented systemic fungal infections. Ketoconazole was more often used for prophylaxis of fungal infections and treatment of oral and esophageal infections.

Use of antifungal therapy in hospitalized patients. II. Results after the marketing of fluconazole.

OBJECTIVE: To evaluate the prescribing patterns of antifungal agents in the hospital setting after the introduction of fluconazole, a new broad-spectrum bis-triazole antifungal agent. Also compared are the prescribing patterns of antifungal agents prior to (phase I) and following (phase II) fluconazole marketing. DESIGN: A prospective cohort of hospitalized patients prescribed topical or systemic antifungal agents. Data were collected from December 1990 to April 1991. SETTING: Fifty-seven hospitals ranging in size from 100 to more than 500 beds. Sixty-three percent are affiliated with medical schools. PATIENTS: Participating pharmacists consecutively identified 15 patients receiving systemic antifungal therapy and 5 patients receiving topical antifungal therapy. INTERVENTIONS: Observational data on patient antifungal therapy, risk factors for fungal infections, comorbidities, concurrent medications, and culture data were collected. MEASURES: Differences in prescribing patterns before and after the marketing of fluconazole were assessed using t-tests and chi-square tests. RESULTS: Of 818 patients studied, 615 (75.2 percent) received systemic antifungal therapy. Five hundred forty-six patients received a single antifungal agent; 348 (63.7 percent) received fluconazole, 105 (19.2 percent) received ketoconazole, 92 (16.8 percent) received amphotericin B, and 1 (0.2 percent) received flucytosine. Sixty-nine patients received two or more systemic agents either concurrently or consecutively. The use of parenteral amphotericin B, alone or in combination with flucytosine and/or an azole, declined from 56.8 percent in the phase I study to 24.2 percent in the current study. The use of parenteral therapy also declined from 56.8 to 40.2 percent. Ketoconazole was used in more than 90 percent of the oral and esophageal infections in the phase I study, but its use declined to only 33 percent in this study. Fluconazole was used most frequently across all sites of presumed or documented infections, with the exception of fungemia. Of the presumed or proven systemic or blood infections, amphotericin B was used alone or in combination in 48.4 percent of the patients and fluconazole was used exclusively in 39.0 percent of the patients. Fluconazole was used more often than amphotericin B (22 vs. 3 patients, respectively) for prophylaxis of systemic infections. The overall use of antifungal prophylaxis also increased from the phase I (9.5 percent) to phase II (13.7 percent). CONCLUSIONS: The introduction of fluconazole had a major impact on the prescribing patterns of antifungal therapy. Although amphotericin B remained the preferred agent for treatment of suspected or proven systemic, central nervous system, or blood infections, use of fluconazole for these indications approached nearly 40 percent. Further studies are needed to address the role of fluconazole in the prophylaxis and treatment of systemic mycoses.

The initial clinical experience of 1819 physicians in maintaining anesthesia with propofol: characteristics associated with prolonged time to awakening.

In October 1989, propofol underwent Phase IV Food and Drug Administration testing that involved 25,981 patients, 1722 institutions, and 1819 anesthesiologists. Participants were 18-80 yr of age and ASA physical status I-III; they could not have a continuing pregnancy or prior adverse anesthetic experience. Anesthesiologists completed detailed forms to describe their use of propofol in this three-step study: propofol for induction only (Step 1), for induction and then maintenance by intermittent bolus injection (Step 2), or for continuous infusion (Step 3). In early 1992, our group of anesthesiologists and epidemiologists analyzed the resulting data base. We evaluated data from 14,882 patients (8095 given bolus injections and 6787 given continuous infusion) to determine factors predicting prolonged time (> 15 min after cessation of all anesthesia) to awakening, one measure of recovery from anesthesia. The incidence of prolonged awakening was 6.8% (1016 patients); the median and mean (+/- SD) times to awakening were, respectively, 5 min and 7.2 +/- 7.3 min. The following variables were associated (P < 0.05) with prolonged awakening from propofol maintenance anesthesia: a total dose of propofol > 8 mg/kg, male gender, endotracheal intubation, age > 65 yr, abdominal surgery, continuous infusion of propofol, and concomitant use of isoflurane or benzodiazepines. These results support the clinical impression that recovery from propofol anesthesia is remarkably rapid; although the vast majority of physicians participating in this study were using propofol for maintenance for the first time, only 6.8% of patients had awakening times exceeding 15 min.

Effects on recovery when isoflurane is used to supplement propofol-nitrous oxide anesthesia.

During propofol-nitrous oxide (N2O) anesthesia, volatile anesthetics are frequently administered to treat signs of inadequate anesthesia and to decrease the possibility of intraoperative awareness. Because the clinical effects of this combination have not been examined rigorously, we used data from the 1989-90 Phase IV clinical trial with propofol to evaluate recovery from propofol-N2O anesthesia with and without supplementation with isoflurane. In this study involving 15,806 patients at 1722 institutions, propofol was administered for induction and maintenance of anesthesia with N2O for procedures lasting less than 60 min. At the discretion of the anesthesiologist, volatile anesthetics were administered as needed during maintenance of anesthesia (the incidence of use of inhaled anesthetics was 14.7% for isoflurane, 2.2% for enflurane, and 0.2% for halothane). Other intraoperative medications included opioid analgesics, muscle relaxants, and anticholinergic drugs. The present study concerns the subset of 7796 patients given propofol-N2O maintenance anesthesia (intermittent bolus or continuous infusion) with or without isoflurane supplementation for procedures lasting less than 60 min. Isoflurane was used more frequently for procedures lasting 30-60 min than for those less than 30 min. Nevertheless, the maintenance dose of propofol was significantly (P < 0.05) less with isoflurane (178 vs 235 mg). Adjunctive use of isoflurane prolonged the time to awakening and to becoming oriented, but discharge times were similar for the two groups. The incidence of postoperative nausea, vomiting, recall, and excitement did not differ between the two groups. We conclude that the addition of isoflurane to a propofol-N2O anesthetic does not alter recovery from anesthesia.

Hemodynamic effects of propofol: data from over 25,000 patients.

To investigate clinically important hypotension and bradycardia after induction of anesthesia with propofol, we analyzed data from a Phase IV stepwise study involving 25,981 patients, 1722 institutions, and 1819 anesthesiologists. In Step 1, propofol was used for induction only. In Step 2, propofol was used for induction and then maintenance by intermittent injection. In Step 3, an induction dose was followed by a maintenance infusion. Participants were to be 18-80 yr of age and ASA physical status I-III; they could not have a continuing pregnancy or prior adverse anesthetic experience. Detailed data on demographic, perioperative, and outcome variables were recorded on data collection forms. The overall incidence of hypotension (systolic blood pressure < 90 mm Hg) was 15.7%; 77% of the episodes were recorded within 10 min of induction of anesthesia with propofol. Bradycardia (heart rate < 50 beats/min) occurred in 4.8% of patients, with 42% of the episodes in the first 10 min. Only 1.3% of patients had both hypotension and bradycardia. The incidence of hypotension was significantly higher for the elderly, females, Caucasians, those undergoing abdominal and integumentary procedures, and those given propofol with opioids, benzodiazepines, or propranolol. Bradycardia was significantly more common when propofol was combined with opioids or chronically taken beta-adrenergic receptor-blocking drugs. Bradycardia and hypotension were not commonly associated. Giving this new drug by protocol, even inexperienced anesthesiologists incurred few adverse hemodynamic changes. Hemodynamic changes were transient and rarely (< 0.2%) required drug therapy. Cardiovascular changes and drug interactions were predictable and manageable based on knowledge of the pharmacology of propofol.

Adverse events in a multicenter phase IV study of propofol: evaluation by anesthesiologists and postanesthesia care unit nurses.

Phase II and III studies are tightly controlled trials investigating adverse effects before government approval of a new drug. However, because postapproval Phase IV studies involve a much larger and more complex population, the true nature of adverse effects can be seen. We analyzed Phase IV data for the new drug propofol with regard to the incidence of adverse events, and evaluations of such events by anesthesiologists versus postanesthesia care unit (PACU) nurses. Data pertained to 25,981 patients, 1722 institutions, and 1819 anesthesiologists giving propofol in three anesthetic regimens. Inclusion criteria were liberal: age, 18-80 yr; ASA physical status I-III; no continuing pregnancy; and no prior adverse anesthetic experience. Anesthesiologists and PACU nurses used data collection forms to record demographic, perioperative, and outcome variables; to evaluate recovery (excellent, good, or poor); and to describe adverse events. Adverse events were reported for 2813 patients (10.8%); the most common events were pain on injection (5.2%), hypotension (1.1%), nausea/vomiting (1.9%), and excitement (1.3%). The incidences of pain on injection and nausea/vomiting were approximately one-half and one-fifth, respectively, the values reported in earlier studies. Six hundred thirty-three patients (2.4%) had a "poor" recovery according to one or both of the evaluators (the anesthesiologist or PACU nurse). The PACU nurse was more influenced by nausea, vomiting, or postoperative pain; and the anesthesiologist was more influenced by postoperative confusion or delayed emergence from anesthesia. For only 0.6% of patients did both evaluators rate recovery as poor. Anesthesiologists gave more weight to intraoperative adverse events, and nurses to postoperative events.(ABSTRACT TRUNCATED AT 250 WORDS)

How do anesthesiologists select patients when introducing a new drug into practice?

As part of the marketing strategy for the anesthetic drug propofol (Diprivan), Stuart Pharmaceuticals began a Phase IV postmarketing study soon after the drug received Food and Drug Administration approval in 1989. We used data from this study to test the hypothesis that anesthesiologists would initially use propofol for young, relatively healthy patients and then, with experience, for older, sicker patients. The Phase IV study involved 1722 institutions, 1819 anesthesiologists, and 25,981 patients. The study incorporated three sequential steps, each to be tested in five patients. In Step 1, propofol was used for induction only; in Step 2, for induction and maintenance of anesthesia by intermittent injection; and, in Step 3, for induction and maintenance by continuous infusion. Inclusion criteria were age 18-80 yr and ASA physical status I-III. Exclusion criteria were continuing pregnancy and a previous adverse anesthetic experience. Physicians used standardized data collection forms to voluntarily compile detailed demographic, perioperative, and outcome variables for patients. Data were then evaluated by an independent, multicenter team of seven anesthesiologists and three epidemiologists to determine whether the first two patients selected to participate in each step (Patients 1 and 2, 6 and 7, and 11 and 12) were less sick, younger, or undergoing less invasive or shorter procedures than patients enrolled later in the same steps (Patients 4 and 5, 9 and 10, and 14 and 15). Physicians gave propofol first to patients with fewer concurrent diseases than are found in the general population (10% were hypertensive versus 16%; 3% were diabetic versus 10%).(ABSTRACT TRUNCATED AT 250 WORDS)

Phase IV study of propofol: validation of the data set.

In 1989-1990, Stuart Pharmaceuticals conducted a Phase IV study of propofol on over 26,000 patients, later making the large data base available to a team of epidemiologists and anesthesiologists for analysis. We now describe the process of verifying the data to provide a sound basis for further analyses. Original data were collected by 1819 physicians at 1761 hospitals. In that study, anesthesia was induced by bolus injection of propofol and was maintained by inhaled drug and N2O-O2 (Step 1), or by propofol (either intermittent bolus injection [Step 2] or continuous infusion [Step 3]) and N2O-O2. Forty-six recorded variables described history, physical examination, course and quality of anesthesia and recovery, and adverse events. Data were scrutinized for inaccuracy or bias regarding adverse events, completeness of data, data entry, and violations of the study protocol. The initial data set pertained to 26,841 patients (10,698, Step 1; 8886, Step 2; and 7257, Step 3). Because we excluded data if 25% of the items were missing from the data set, 3.2% of the case reports were eliminated: the final data set used for subsequent analyses contained 25,981 patients (10,184, Step 1; 8672, Step 2; and 7125, Step 3). Inaccuracy of data entry was not excessive, and violations of study protocol were less frequent than in similar studies. The nature and frequency of adverse events were similar to those reported in Phase II and III clinical trials of propofol. Analysis showed that missing data occurred randomly and did not introduce obvious bias. We conclude that the data set was valid and most likely represents perioperative events occurring in similar patients; that Phase IV studies can be valuable because of the range of patients studied and the ability to detect even rare events; and that future Phase IV studies could be improved by more efficient design of data collection forms for both hypotheses to be tested and the entry of data onto forms.

The role of pharmacoepidemiology research in postmarketing surveillance and anesthesia practice/critical care medicine.

Despite tremendous efforts to ensure the safety and effectiveness of newly marketed medications, a number of these have had significant problems after introduction of the drug to the market. Such problems highlight the practical limitations of clinical trials performed to obtain FDA approval for marketing. Pharmacoepidemiology research methodologies provide a powerful mechanism for exploring the determinants of drug safety and effectiveness in broad-based populations and can serve as a scientific foundation for outcome research. Using these methodologies, largescale postmarketing surveillance studies similar to the type described in the accompanying articles would constitute an important way of confirming and identifying the determinants of drug safety and effectiveness in large, diverse patient populations.

Capability of hospital computer systems in performing drug-use evaluations and adverse drug event monitoring.

A survey to determine the extent of computerization in key areas of hospitals, the information being collected in the databases, and the capabilities of the computer systems for performing adverse drug event monitoring and drug-use evaluations was conducted. The questionnaire was distributed to clinical pharmacists in the 500 hospitals composing the Drug Surveillance Network. In the majority of the 166 responding hospitals (> 85%), the pharmacy department, clinical chemistry and hematology laboratories, patient admissions, and microbiology laboratory were computerized for data acquisition and management. The medical records and purchasing departments were computerized in a smaller proportion of hospitals (75% and 74%, respectively). In the majority of hospitals with a computerized pharmacy department (> 78%), there was ready access to computer databases in other departments, but simultaneous querying of multiple databases was possible in only 30%. Patients could be identified according to diagnosis in 82% of the hospitals and according to medication received in 83%. More than 85% of responding hospitals had implemented spontaneous reporting systems for the identification of adverse drug events. Computers are widely used in hospitals participating in the Drug Surveillance Network, but a substantial effort is necessary to make these resources more useful and to standardize processes so that data may be pooled across institutions to deal with important public health concerns.

The clinical evaluation of new drugs for sepsis. A prospective study design based on survival analysis.

OBJECTIVE: To develop a survival model and severity assessment method to estimate the 28-day mortality risk for patients with sepsis syndrome entering phase 2 and 3 drug evaluations. DESIGN: Retrospective analysis of intensive care unit admissions with sepsis syndrome by means of log-normal regression to identify risk factors for 28-day mortality. Prospective application of the model to patients with gram-negative infection meeting sepsis syndrome criteria from separate data collection (validation group). PATIENTS: A total of 58,737 intensive care unit admissions at 107 hospitals in the United States and Western Europe screened to yield 1195 patients meeting entry criteria for the sepsis syndrome study for the original model; 295 hospitalized patients with gram-negative infection meeting criteria for sepsis syndrome for validation. MAIN OUTCOME MEASURES: Survival time and mortality at 28 days after fulfillment of the sepsis syndrome criteria. RESULTS: Acute physiologic abnormalities were the most important prognostic factors influencing outcome (82% of total chi 2). Specific disease resulting in intensive care unit admission and the time the patient was in the hospital and intensive care unit before qualification were also independent risks, as were age and a clinical history of cirrhosis. The model's overall classification accuracy was a Somers' Dyx of .52 (rank correlation between predicted risk and 28-day mortality) (receiver operating characteristic area, 0.76), with equal accuracy (Dyx = .59; receiver operating characteristic area, 0.80) in the independent group of patients. CONCLUSIONS: We created an accurate independent estimate for 28-day mortality risk for patients with sepsis syndrome (severe sepsis). This estimate could improve the evaluation of new drugs by investigating whether the drug's benefit varies by patient risk and then determining the amount of benefit for individual patients.