Applying control charts to quality improvement.

This article reviews the definitions and principles that should guide the use of control charts in healthcare quality. Several examples from the literature are used to illustrate significant problems and issues in control chart construction.

A comparison of three measures of disease progression in ALS.

Effective clinical trials in neuromuscular research require accurate and sensitive methods to quantitate disease progression. The purpose of this study was to concurrently compare manual muscle testing (MMT), maximal voluntary isometric contraction (MVIC), and a functional scale (the ALS Score). Twenty patients with ALS were tested ten times at monthly intervals using each of the three methods. High inter-rater reliability for each of the three methods was demonstrated. All three methods demonstrated a decline over the testing period. However, MMT and the ALS Score were less sensitive to early change. A wide variation of percent of normal MVIC was demonstrated within each MMT grade and extensive overlap between MMT grades was observed. MVIC is the method of choice in studying patients in the early stages of ALS and in studies conducted over a short period of time.

Perspective: comparing two-sample tests and control charts.

For the task of monitoring a process, detecting change, and making correct attributions on a continuous basis, there is no better tool than the control chart, particularly where the data set is large and permits subgrouping. By using this technique, staff will get quicker answers or will have a smaller data requirement, and will arrive at more valid conclusions than with the two-sample method. (The two-sample approach is typically used when the data are not produced or collected in sequence is unknown. This, however, may reflect a weakness in the data collection process.) In addition, we believe that the control chart approach is as simple as the two-sample approach. Once the control limits and zones are established, the plot of points can be maintained by personnel who need to do no more than plot a rate on a weekly (or a daily) basis. The control chart rules properly focus laboratory and QA personnel on change. Control chart procedures are statistical tests of special cause and need no further significance tests. Those using control charts need to be aware of three important points: 1. Control limits are not standards or thresholds and cannot be derived from these concepts. Control limits are a statement about the current natural variation in the process. 2. The type of chart to be used should depend upon the type of data being evaluated. (Finison et al., 1993).(ABSTRACT TRUNCATED AT 250 WORDS)

The use of control charts to improve healthcare quality.

Control charts are a basic tool for understanding variation in all healthcare processes. Control chart limits are not standards; rather, they divided variation into special and common cause, each of which requires a different management response. Each type of data--variables, count defect, or defectives data--requires a different type of chart (e.g., X bar and R, X bar and S; C or U; P or NP). Individual observations where events are infrequent can be charted on an individual's chart with moving range limits or by several alternative methods discussed in this article.

Use of electrophysiologic tests to measure disease progression in ALS therapeutic trials.

A battery of electrophysiologic tests was developed to assess the relative degree of lower and upper motor neuron (spasticity) deficit in a group of ALS patients enrolled in a therapeutic trial. Test results were correlated with strength in the tibialis anterior muscle as determined by measurement of maximum voluntary isometric contraction (MVIC), using strain gauge tensiometers, and were also correlated with a clinical spasticity rating scale. Patients were tested every 6 to 8 weeks over more than 1 year. Compound muscle action potential amplitude (CMAPa) from tibialis anterior correlated best with MVIC and also showed a strong linear correlation with time, as did MVIC. Other tests correlated poorly with MVIC on the average, although individual patients did show high correlations. In those patients where correlation between CMAPa and MVIC was low, MVIC did not show a high linear correlation with time and was also highly variable. This study suggests that the addition of CMAPa should be considered in ALS therapeutic trials if MVIC is not available. In addition, CMAPa can be useful in study samples where MVIC deterioration is not linear over time or is highly variable.

Use of composite scores (megascores) to measure deficit in amyotrophic lateral sclerosis.

The Tufts Quantitative Neuromuscular Exam (TQNE) consists of 28 items that were designed to measure voluntary motor deficit in amyotrophic lateral sclerosis (ALS) and related diseases. Individual raw data were converted to Z scores for standardization and then grouped into five megascores with statistical and clinical relevance. The derived megascores were Mega 1, pulmonary function; Mega 2, bulbar function; Mega 3, timed hand activities; Mega 4, isometric arm strength; and Mega 5, isometric leg strength. Megascores should enhance the usefulness of testing in therapeutic trials and in analyzing the natural history of ALS and related diseases.

Quantitative assessment of neuromuscular deficit in ALS.

Several criteria must be met in developing a test battery forr ALS. First, the test items should be designed to answer the specific questions being asked. If questions involve the amount and rate of deterioration, then the test items should generate interval data and show evidence that they reflect change in the disease itself. Second, the reliability of the test items should be carefully determined. Are the variances due to testing errors such as inconsistent positioning or changes in verbal instructions; or do they reflect human performance variation? Third, the test items must be sensitive to small changes, and one must be able to test the very strong as well as the severely debilitated patient. The TQNE was designed to measure degree of disease progression in ALS. The major portion of the protocol consists of isometric force measurement using an electronic strain gauge. Other quantitative segments include tests of pulmonary function, speech rates, and timed activities. Using z-score transformations, we have summarized these test items into five megascores: pulmonary function, bulbar function, timed hand activities, arm strength, and leg strength. Megascores have been used to characterize the natural history of the disease and to judge therapeutic efficacy. Quantitative methods to measure deficit in neuromuscular disease lag far behind other technologic advances in medical research. Future advances in the measurement of neuromuscular deficit depend on the commitment of researchers to invest the time and resources necessary to develop more sensitive and reliable testing techniques.

Intrathecal administration of natural human interferon alpha in amyotrophic lateral sclerosis.

Ten patients with amyotrophic lateral sclerosis were given intrathecal injections of natural interferon alpha, 1 million units weekly for 7 to 24 weeks. Six patients completed the trial. Four voluntarily withdrew after 7 to 13 injections. The slopes of deterioration for 40 quantitative tests of neuromuscular function for the control and treatment periods were compared by paired t test in the six patients who completed the trial and in the patient who withdrew after 13 injections. No significant differences were found. The patients tolerated treatment well. The CSF reaction was modest and spontaneously reversible. Indomethacin and ibuprofen blocked interferon side effects.

Blood pressure reductions: correcting for regression to the mean.

Reductions in high blood pressure (BP) from participating in screening and treatment programs are often assessed by comparing BP measurements before and after participation. The interpretation of such changes in measured blood pressure is confounded by the tendency of high pressures to decline as a result of a statistical artifact--regression to the mean. The problem arises whenever baseline measurements are used both for selection of participants and for comparisons with pressures obtained later. We developed a statistical model which predicts the average decline due to regression for participants in a screening or treatment program. This regression effect must be subtracted from the observed reduction in BP (the difference between baseline and later measurements) to obtain the average net reduction in BP from the program. The regression effect is estimated as the product of two factors. The first factor is the proportion of the variance in the baseline (preprogram) measurement due to measurement error and the short-term variation (e.g., 0.24 for two replications averaged). The second factor is the difference between the mean baseline pressure of full participants and that of the underlying population of potential participants. The model was first illustrated with successive BP measurements from community screening programs, where the "program" was only remeasurement. The mean observed decline in diastolic BP between screens for 145 persons with elevated baseline BP was 7 mm Hg. After adjustment for regression to the mean, the net decline between screens was estimated to be 2 mm Hg. This decline is apparently due to the pressor effect, or stress of screening, and agrees with findings from other studies. Next the model was applied to the treatment phase of the Hypertension Detection and Follow-up Program. Overall, net reductions predicted by the model agree with those from independent measurements to within 0.1 mm Hg. The findings indicate the one can compute net reductions in BP from before-and-after comparisons in screening and treatment programs with reasonable accuracy, and these net reductions are generally much smaller than the crude BP declines.