Randomized clinical trial of a quality improvement intervention in nursing homes.

PURPOSE: The purpose of the study was to determine if simply providing nursing facilities with comparative quality performance information and education about quality improvement would improve clinical practices and subsequently improve resident outcomes, or if a stronger intervention, expert clinical consultation with nursing facility staff, is needed. DESIGN AND METHODS: Nursing facilities (n = 113) were randomly assigned to one of three groups: workshop and feedback reports only, workshop and feedback reports with clinical consultation, and control. Minimum Data Set (MDS) Quality Indicator (QI) feedback reports were prepared and sent quarterly to each facility in intervention groups for a year. Clinical consultation by a gerontological clinical nurse specialist (GCNS) was offered to those in the second group. RESULTS: With the exception of MDS QI 27 (little or no activity), no significant differences in resident assessment measures were detected between the groups of facilities. However, outcomes of residents in nursing homes that actually took advantage of the clinical consultation of the GCNS demonstrated trends in improvements in QIs measuring falls, behavioral symptoms, little or no activity, and pressure ulcers (overall and for low-risk residents). IMPLICATIONS: Simply providing comparative performance feedback is not enough to improve resident outcomes. It appears that only those nursing homes that sought the additional intensive support of the GCNS were able to effect enough change in clinical practice to improve resident outcomes significantly.

Initial field testing of an instrument to measure: observable indicators of nursing home care quality.

The "Observable Indicators of Nursing Home Care Quality" instrument was developed as a new measure of nursing home care quality. The instrument is based on a theoretical model of quality nursing home care grounded in data from provider and consumer focus groups. The instrument was piloted in 10 Missouri nursing homes. Subsequent versions were tested in 109 Missouri and 11 Icelandic nursing homes. Content validity was established using experts. Concurrent and known groups validity was evaluated using Minimum Data Set quality indicators, survey citations, and a process of care measure. Interrater and test-retest reliabilities were calculated as well as coefficient alpha. The "Observable Indicators of Nursing Home Care Quality" instrument is a new measure that can be used by researchers, and potentially by regulators, consumers, or providers, to observe and score specific indicators of quality care following a 20- to 30-minute inspection of a nursing home.

Validation of a method for the determination of zolpidem in human plasma using LC with fluorescence detection.

A sensitive and selective high-performance liquid chromatography (HPLC) method was developed for the determination of zolpidem in human plasma. Zolpidem and the internal standard (trazodone) were extracted from human plasma that had been made basic. The basic sample was loaded onto a conditioned Bond Elut C18 cartridge, rinsed with water and eluted with methanol. Forty microliters were then injected onto the LC system. Separation was achieved on a C18 column (150 x 4.6 mm, 5 microm) with a mobile phase composed of acetonitrile:50 mM potassium phosphate monobasic at pH 6.0 (4:6, v/v). Detection was by fluorescence, with excitation at 254 nm and emission at 400 nm. The retention times of zolpidem and internal standard were approximately 4.7 and 5.3 min, respectively. The LC run time was 8 min. The assay was linear in concentration range 1-400 ng/ml for zolpidem in human plasma. The analysis of quality control samples for zolpidem (3, 30, and 300 ng/ml) demonstrated excellent precision with relative standard deviations (RSD) of 3.7, 4.6, and 3.0%, respectively (n = 18). The method was accurate with all intraday (n = 6) and overall (n = 18) mean concentrations within 5.8% from nominal at all quality control sample concentrations. This method was also performed using a Gilson Aspec XL automated sample processor and autoinjector. The samples were manually fortified with internal standard and made basic. The aspec then performed the solid phase extraction and made injections of the samples onto the LC system. Using the automated procedure for analysis, quality control samples for zolpidem (3, 30, and 300 ng/ml) demonstrated acceptable precision with RSD values of 9.0, 4.9, and 5.1%, respectively (n = 12). The method was accurate with all intracurve (n = 4) and overall (n = 12) mean values being less than 10.8% from nominal at all quality control sample concentrations.

Validation of a method for the determination of (R)-warfarin and (S)-warfarin in human plasma using LC with UV detection.

A sensitive and selective chiral high-performance liquid chromatography (HPLC) method was developed for the determination of (R)-warfarin and (S)-warfarin in human plasma. (R)- and (S)-warfarin and the internal standard (oxybenzone) were extracted from human plasma that had been made acidic with 1 N sulfuric acid into ethyl ether. The ethyl ether layer was removed and evaporated, and the residue was reconstituted in 200 microl of acetonitrile. A 50-microl aliquot was injected onto the HPLC system. Separation was achieved on a beta-cyclodextrin column (250 x 4.6 mm, 5 microm) with a mobile phase composed of acetonitrile:glacial acetic acid:triethylamine (1000:3:2.5, v/v/v). Detection was by ultraviolet absorbance at 320 nm. Late-eluting peaks were diverted from the analytical column by using a beta-cyclodextrin precolumn (50 x 4.6 mm, 5 microm) and a column switching device. The retention times of (R)- and (S)-warfarin and the internal standard were approximately 7.7, 6.9 and 4.0 min, respectively. The run time was 15 min. The assay was linear in concentration ranges of 12.5-2500 ng/ml for (R)- and (S)-warfarin in human plasma. The analysis of quality control samples for (R)- and (S)-warfarin (25.0, 400 and 2000 ng/ml) demonstrated excellent precision with relative standard deviations (R.S.D.) for (R)-warfarin of 10.9, 2.8, and 2.8%, respectively (n = 18), and for (S)-warfarin of 7.0, 2.4 and 2.6%, respectively (n = 18). The method was accurate with all overall (n = 18) mean concentrations being less than 6.0% from nominal at all quality control sample concentrations.

Nursing home care quality: a multidimensional theoretical model integrating the views of consumers and providers.

This exploratory study was undertaken to discover the defining dimensions of nursing home care quality from the viewpoint of consumers of nursing home care. Eleven focus groups were conducted in five Missouri communities. The seven dimensions of the consumer multidimensional model of nursing home care quality are: staff, care, family involvement, communication, environment, home, and cost. The views of consumers and families are compared with the results of a previous study of providers of nursing home services. An integrated, multidimensional theoretical model is presented for testing and evaluation. An instrument based on the model is being tested to observe and score the dimensions of nursing home care quality.

Improvements in detection sensitivity for the determination of ivermectin in plasma using chromatographic techniques and laser-induced fluorescence detection with automated derivatization.

Improvement in detection sensitivity for the analysis of ivermectin was observed through utilization of laser-induced fluorescence detection and by manipulation of chromatographic conditions. Gradient elution used in combination with narrow-bore chromatography and conventional fluorescence detection resulted in a limit of quantitation for the major homologue of ivermectin of 0.01 ng/ml in dog plasma. Laser-induced fluorescence detection with isocratic chromatographic conditions also resulted in a limit of quantitation of 0.01 ng/ml in dog plasma, which is a six-fold improvement over previously reported methods. Introduction of an automated procedure for the derivatization and injection of samples reduced the amount of sample handling, eliminated the potential for analyte/internal standard degradation and contributed to the overall ease of analysis.

A minicomputer-based perinatal/neonatal telecommunications network.

A minicomputer-based telecommunications network has been developed to link all of the state's nine tertiary neonatal intensive care units in order to provide information on availability of neonatal and maternal beds. The information is updated automatically on computer terminals located in each neonatal intensive care unit every two hours by a complete system poll (350 polls per month). This information has been found to be accurate 95% of the time. Through a computer simulation, this network is shown to provide a significant reduction (P less than .0005) in the number of referral telephone calls, and thus the time, required to locate a bed within the state for obstetric or pediatric patients needing transfer. In addition, this network has provided a data base to aid in local, regional, and statewide planning for perinatal facilities. A 6-month summary of the data indicated that for 16% of the system polls, less than two neonatal intensive care units could accept outside referrals, demonstrating frequent severe statewide restrictions for care of high-risk patients. The total operating budget for the most recent fiscal year was less than +25,000, which is far less than that of manual 24-hour central telephone operator systems providing similar services. This telecommunications network is a cost-effective answer to the common problem of limited regional perinatal resources.