Analysis of pain management in critically ill patients.

We analyzed the adequacy of pain control for 17 trauma patients during the initial part of their stay in the intensive care unit, and assessed reasons for inadequate analgesia, if it occurred. Patients, and physicians, and nurses were interviewed. A verbal pain intensity scale was used to determine whether patients received adequate analgesia. Patients were asked if the pain hindered their activities, and whether they requested pain medication from their caregivers. Caregivers were questioned whether patients received adequate analgesia. Prescribed morphine regimens and the amount of narcotic administered were analyzed. Twenty-seven percent of patients rated pain intensity as moderate and 47% as severe. Ninety-five percent of housestaff and 81% of nurses reported the patients received adequate pain control. Forty-seven percent of the patients who had moderate or severe pain asked their physician for more pain medication, and 65% asked the nurse. Thirteen residents did not order a larger dose of morphine due to concern about respiratory depression or hypotension. Morphine dosages ranged from 1-8 mg intravenously every 1-2 hours as necessary. Nurses administered less than the maximum amount ordered 58% of the time. The mean dosing interval was 2.3 hours. Barriers to adequate pain management were disparity in the perception of pain between patients and caregivers; patients not requesting more analgesia despite despite the presence of moderate to severe pain; and physician and nurse concerns about patients' adverse physiologic response to increased dosages.

Identification of patient-controlled analgesia overdoses in hospitalized patients: a computerized method of monitoring adverse events.

OBJECTIVE: To describe and validate a computer-based quality assurance method that detects narcotic overdoses associated with patient-controlled analgesia (PCA) use. SETTING: Two acute care teaching hospitals. PATIENTS: 4669 patients who received PCA. INTERVENTIONS: The following patient lists were obtained during a two-year period from both hospital information systems: those who received PCA and (1) received naloxone, a narcotic antagonist, (2) were transferred to an intensive care unit, (3) had a cardiac or respiratory arrest, or (4) died. Possible overdoses were defined as patients who appeared on the PCA list and one of the other lists. Charts were reviewed if the patient's name appeared on the PCA and one of the other lists. Patients were judged to have experienced a narcotic overdose if there was an immediate improvement in blood pressure, respiratory rate, or mental status after the administration of naloxone. RESULTS: The search strategy identified 294 possible overdoses in 1499 patients who received PCA. Ten charts were unavailable for review. An actual overdose occurred in 11 patients. The accuracy of the new method was compared with that of the hospitals' present reporting methods. Eleven overdoses were identified by the computer search, but only 6 overdoses were identified in incident and adverse drug reaction reports. CONCLUSIONS: The systematic computer search identified almost twice as many adverse incidents than were reported by the traditional hospital methods.

Difficulties in diagnosing narcotic overdoses in hospitalized patients.

OBJECTIVE: To describe the clinical presentation of narcotic overdose in hospitalized patients and to differentiate this circumstance from other conditions often misdiagnosed as overdose. DESIGN: Case series. SETTING: Two acute-care teaching hospitals. PATIENTS: Forty-three hospitalized patients who received naloxone for a clinically suspected narcotic overdose. INTERVENTIONS: Two investigators independently evaluated each incident to determine whether the patient had a narcotic overdose. The patients were judged to have had an overdose if caregivers documented an immediate improvement in mental status, respiratory rate, or blood pressure after naloxone administration. MEASUREMENTS: The clinical presentation of a narcotic overdose in hospitalized patients was defined. Conditions misdiagnosed as an overdose were determined. MAIN RESULTS: Symptoms improved rapidly with the administration of naloxone in 28 incidents (65 percent) and were designated overdose. In 15 other instances there was no improvement in symptoms; these patients were designated nonoverdose. Only half of the overdose patients had a respiratory rate < 8 breaths/min immediately prior to naloxone administration. Only two of the overdose patients had the classic triad of symptoms (respiratory depression, coma, and pinpoint pupils). Other overdose patients had only one or two of the classic signs. The clinical presentation of narcotic overdoses in hospitalized patients did not include respiratory depression, hypotension, or coma in the majority of patients. All overdose patients showed a decrease in mental status. The majority of nonoverdose patients had pulmonary conditions that were misdiagnosed as a narcotic overdose. CONCLUSIONS: Narcotic overdoses in hospitalized patients seldom fit the classic description. The lack of respiratory depression does not mean the absence of a narcotic overdose. Patients who receive narcotics and develop a significant decrease in mental status should be evaluated for a possible overdose. Pulmonary, neurologic, cardiovascular, and electrolyte abnormalities often are misdiagnosed as a narcotic overdose in hospitalized patients.

Pharmacokinetic evaluation of a new oral cyclosporine formulation.

STUDY OBJECTIVE: To compare the pharmacokinetics of a new oral cyclosporine preparation with those of cyclosporine solution diluted in Isocal and the intravenous formulation. DESIGN: Randomized, crossover trial. SETTING: Tertiary care referral center. PATIENTS: Seven pediatric liver transplant recipients who were receiving oral cyclosporine as part of their immunosuppressive regimen. All patients completed the study. INTERVENTIONS: Pharmacokinetic studies were performed with the intravenous and oral dosage forms. Patients received one dose of intravenous cyclosporine, and then were randomized to receive their usual oral cyclosporine dose incorporated into a chocolate wafer or mixed with Isocal. After a minimum of 3 days, the alternative preparation was administered. Serial cyclosporine blood samples were collected at predetermined intervals for 12 hours after the third dose for each regimen. Concentrations were determined by high-performance liquid chromatography. The data for the three dosage forms were fit simultaneously with a two-compartment model. MEASUREMENTS AND MAIN RESULTS: No difference was seen in F, ka, Cmax, and tmax between the two oral cyclosporine preparations (p > 0.05). No new rejection episodes occurred during the study period. CONCLUSIONS: We conclude there is no difference in the bioavailability of the oral solution and the chocolate formulation. We believe the new preparation may increase patient compliance and ensure administration of a complete dose compared with the currently marketed solution.

Glucose response to abrupt initiation and discontinuation of total parenteral nutrition.

Plasma glucose was studied during the initiation of total parenteral nutrition (TPN) and the discontinuation of TPN without a tapering schedule. Blood was sampled every 5 minutes for 2 hours after the start of TPN and 1 week later as TPN was discontinued. A total of 14 initiations and 14 discontinuations were studied in 18 patients. Severity of illness in patients ranged from stable condition postoperatively to multiple-system failure; six patients had diabetes mellitus. The TPN solution was a 3:1 admixture that provided a caloric intake equal to 1.2 times the resting energy expenditure, with 40% fat and 60% carbohydrate calories. An average of 1963 kcal was provided per day (340 g of glucose, 79 g of fat). During the initiation phase, the mean increase in plasma glucose was 60 mg/dL. The increase for diabetic patients was 79 +/- 14 mg/dL compared with 52 +/- 23 mg/dL for the nondiabetics. During the discontinuation phase, the mean plasma glucose decreased 40 +/- 20 mg/dL; two patients with high concentrations of regular insulin (50 and 100 units) showed an increase in plasma glucose when the TPN was stopped. Plasma glucose returned to the preinfusion baseline after discontinuation. During both initiation and discontinuation, plasma glucose showed little change after the first 60 minutes. No clinical symptoms of hypoglycemia were observed. In conclusion, TPN as a 3:1 admixture can be safely started as full nutrition support and stopped abruptly without a tapering schedule. Plasma glucose response is rapid, predictable, and mostly complete within 60 minutes.

The substance abuser and home intravenous therapy: above all else, do no harm.

Home care therapy is being challenged by changes in patient populations and technologic advances. The selection of appropriate candidates for home intravenous therapy is a critical issue faced by health care professionals. This process is more complex when the patient has a history of intravenous drug abuse. The issues concern patient compliance, safety, ethics, and legal responsibilities. Safe care depends on the ability of the patient to demonstrate a predetermined level of competence with catheter use. The potential use of illicit drugs may influence the ability of the patient to be compliant. Ethical principles of the patient's autonomy and free choice are weighed against the health professional's sense of beneficence. Legal guidelines stress informed consent, standards of care, and adequate documentation. An exploration of each of these factors outlines the potential risks and benefits and provides a basis for making clinical judgments.

Narcotic use in the hospital: reasonably safe?

OBJECTIVE: To determine the causes and frequency of overdoses associated with the administration of opioid analgesics in hospitalized patients. DESIGN: Case series. SETTING: Two acute care teaching hospitals. PATIENTS: Eighty-one hospitalized patients who received naloxone for a clinically suspected narcotic overdose. INTERVENTIONS: Three investigators reviewed each patient who received naloxone during a 12-month period. The patients were judged to have a narcotic overdose if caregivers documented an immediate improvement in mental status, respiratory rate, or blood pressure after naloxone administration. MAIN OUTCOME MEASURES: The number and causes of narcotic overdoses were determined. The frequency of morphine and meperidine overdoses was calculated. The number of incidents reported using incident or adverse drug reaction reports or the appropriate International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code. RESULTS: In the 22 overdoses that occurred, 14 (64 percent) were caused by medication prescribing, compounding, or administration errors and potentially were preventable. The remaining eight patients experienced an overdose despite receiving appropriate amounts of opioids. The frequency of overdoses was 0.4 and 0.2 percent of total patients receiving morphine or meperidine, respectively, at the two hospitals. Nonreporting of these narcotic overdoses was frequent. In one hospital, 1 incident report and 3 adverse drug reactions were reported for 17 overdoses. At the second hospital, 1 incident report and 1 adverse drug reaction were reported for 6 overdoses. None of the patient charts included an ICD-9-CM code that documented the problem. CONCLUSIONS: The causes of overdoses are not limited to prescribing and administration errors. Some patients, despite proper execution of appropriate orders, develop a narcotic overdose. Caregivers must be aware of this problem and monitor patients for a decrease in mental status and respiratory rate. In addition, we conclude that an important number of hospitalized patients develop an overdose even though the frequency is low related to the number of patients receiving narcotics.

Effect of individualized pharmacokinetic dosing on patient outcome.

OBJECTIVE: To study the effect of individualized pharmacokinetic dosing of aminoglycosides on patient outcome. DESIGN: Prospective, randomized study. SETTING: Tertiary care hospital. PATIENTS: Ninety-five patients with documented Gram-negative infections received 97 courses of aminoglycoside therapy. INTERVENTIONS: Patients were randomized between pharmacokinetic dose adjustment and monitoring or traditional physician-directed techniques. Patients were stratified by severity of underlying illness before randomization. MEASUREMENT AND MAIN RESULTS: Sixty-two courses of treatment were satisfactorily completed. Patients in the severely ill group (eight kinetic, eight traditional) had significantly (p less than .05) better survival (7 kinetic, 3 traditional) when managed with pharmacokinetic consultation. The kinetic arm received greater doses (156 +/- 59 mg/dose; 2.4 +/- 0.6 mg/kg) than the traditional arm (81 +/- 27 mg/dose; 1.5 +/- 0.6 mg/kg) (p less than .001). In addition, the dose per day (mg/kg) was greater in the kinetic arm (4.1 +/- 1.5) than the traditional arm (3.2 +/- 1.3) (p less than .001). The improved survival was achieved by attaining therapeutic peak serum concentrations earlier in the course of the infection and by administering more total aminoglycoside without increasing toxicity. CONCLUSIONS: We conclude that pharmacokinetic management of aminoglycoside dosing may improve the outcome of severely ill patients.

Gastric secretion pH measurement: what you see is not what you get!

We evaluated the accuracy of gastric secretion pH measurements as performed in three ICUs. The pH of 275 samples was measured with pH paper using established techniques. The pH of 85 additional samples was determined with a hand-held pH meter. All specimens also were measured using a research laboratory pH meter to learn the true pH. Analyses included mean and SD of the difference between the two measurements, the correlation coefficient (r value), and the concordance correlation coefficient. The pH meter values disagreed significantly with pH paper measurements. Measurements of gastric secretion pH with pH indicator paper do not guide therapy reliably. Inaccurate values derived from pH paper measurements could have resulted in inappropriate treatment in 28% of the samples tested. A portable, battery-powered pH meter accurately reproduced laboratory pH meter measurements and is a reasonable device for clinical use.

Per capita income in breast cancer patients.

The Wisconsin Department of Health and Human Services has identified breast cancer as the most common malignant neoplasm among women in Wisconsin. To determine the potential effectiveness of a program of low-cost screening mammography in reducing cancer morbidity and mortality in individuals of low socioeconomic status, the authors conducted a retrospective analysis of data from tumor registries. Using the tumor registries of two hospitals in Milwaukee, Wisconsin, the records of 323 patients with breast cancer were identified and analyzed for size of tumor at first presentation. These data were correlated with per capita income taken from census block information for the residence of each patient. The data indicate that mean income was lower for patients with more advanced disease. Other variables such as race did not influence tumor classification at initial presentation. Economically disadvantaged women present with more advanced breast cancers than affluent patients. This adverse circumstance may be the result of lack of financial resources and poor dissemination of information. Programs that recognize these problems are likely to be more successful in achieving earlier breast cancer detection.

Prospective comparison of traditional and pharmacokinetic aminoglycoside dosing methods.

Aminoglycoside (gentamicin, tobramycin) dosage regimens and subsequent serum concentrations were compared in 30 patients treated initially using traditional physician-determined methods and then switched to a pharmacokinetic-based treatment program. Patients received more drug during the kinetic phase (median 5 mg/kg) than during the traditional phase (median 3.6 mg/kg) and achieved greater peak serum concentration (5.9 vs. 4.4 micrograms/ml). Seventy-three percent of kinetic peak values but only 27% of traditional peak values exceeded 5.0 micrograms/ml. Trough concentrations were comparable in both phases of study and no nephrotoxicity was observed. This pharmacokinetic-based management program achieved more consistently greater therapeutic peak concentrations and provided more individualized therapy than did physicians. The use of pharmacokinetic consultants may be of benefit in administering safely optimal aminoglycoside therapy.

Epidural morphine by continuous infusion with an external pump for pain management in oncology patients.

Ten patients with advanced malignancies and severe pain were given epidural morphine (EDM) by continuous infusion. The pain had been treated previously with large doses of oral or parenteral narcotics, without success. The pain was disabling in all the cases. The total dose of morphine administered was 10-90 mg per day; median dose was 25 mg. Treatment lasted from 9 to 400 days. Pain relief was complete in seven patients, near complete in two, and moderate in one. Five patients became fully ambulatory. Two patients were treated as outpatients and resumed work activities. No central nervous system or gastrointestinal adverse effects, respiratory depression, or pruritus were noted. One patient developed urinary retention. EDM by continuous infusion produced constant pain relief for prolonged periods. This technique is safe for analgesia in oncology patients and suited for outpatient management.

A quality control analysis of aminoglycoside management.

Aminoglycoside administration practices were evaluated in a teaching hospital using three study methods: a chart review of 40 randomly selected patients receiving aminoglycosides was conducted retrospectively; 93 health care personnel involved in ordering and administering aminoglycosides to patients were interviewed regarding their understanding of aminoglycoside utilization practices; and ten patients having serum peak and trough aminoglycoside determinations were closely monitored for accuracy of dose administration and obtaining blood specimens at appropriate times. The chart review showed that during 15 of 32 evaluable therapy courses no determinations of serum aminoglycoside concentration were obtained. The survey demonstrated that only 24% of the residents actually used the results of peak and trough determinations to adjust dosage regimens. Direct observation of health care personnel disclosed only two of ten instances in which doses were administered and serum concentration specimens obtained with no apparent problems. Most personnel in our hospital were unaware of these pervasive suboptimal or inconsistent practices associated with aminoglycoside administration and interpretation of laboratory results.

Stability of mitomycin admixtures.

The stability of mitomycin in admixtures for continuous intravenous infusion was studied. Mitomycin was reconstituted and diluted to 50 micrograms/mL in polyvinyl chloride minibags containing 5% dextrose injection 50 mL or 0.9% sodium chloride injection 50 mL. Additional mitomycin admixtures were reconstituted with a buffer solution containing monobasic and dibasic sodium phosphate; these were diluted with 5% dextrose injection only. Admixtures were stored at room temperature (27-30 degrees C) and refrigerated temperature (5 degrees C) for 120 days. Mitomycin concentrations in each admixture were tested by high-performance liquid chromatography (HPLC) immediately after admixture and at intervals during storage. Ultraviolet spectra were determined at the same time as HPLC analysis, and the admixtures were visually inspected and tested for pH. Mitomycin concentrations decreased rapidly in the unbuffered admixtures; after 12 hours at room temperature, less than 26% of the drug remained in the dextrose admixture. When the unbuffered admixtures were refrigerated for 12 hours, the mitomycin concentrations decreased 10% in the sodium chloride admixtures and 33% in the dextrose admixtures; after 24 hours, the percentages of drug loss were 23% and 42%, respectively. Mitomycin concentrations in the buffered admixtures showed no substantial decrease during 120 days at 5 degrees C. At room temperature, concentrations decreased 10% after 15 days. When the admixture is buffered to a pH of approximately 7.8, mitomycin is stable in 5% dextrose injection for up to 15 days at room temperature and at least 120 days at 5 degrees C. Unbuffered mitomycin admixtures should not be stored or administered by prolonged i.v. infusion.

Long-term, ambulatory, continuous IV infusion of 5-FU for the treatment of advanced adenocarcinomas.

Twenty-three patients with metastatic adenocarcinomas were treated with long-term, continuous, ambulatory iv infusion of 5-FU. Length of infusion ranged from 54 to 324 days. The usual daily dose was 300 mg/m2. Toxicity was primarily stomatitis. Hand/foot syndrome occurred in 11 patients. Nausea, vomiting, myelosuppression, and alopecia were not observed. Thirteen patients had stomatitis. Eighteen patients had evaluable lesions; eight achieved partial response, five had stable disease, and five had progressive disease. Further studies are necessary to confirm the level of tumor response and survival period of patients treated with this method.