Distribution-based Approach Research Articles

The clinical relevance of changes in the Montgomery–Asberg Depression Rating Scale using the minimum clinically important difference approach

ABSTRACTObjective: To identify the minimal clinically important difference (MCID) for the Montgomery–Asberg Depression Rating Scale (MADRS) in randomised studies of depression, and to cross-validate the estimated MCID.Design and methods: Placebo-treated patients from three similarly-designed, 8-week, double-blind, randomised depression trials with a stable health status between baseline and Week 1 (‘no change’ rating on the Clinical Global Impression–Improvement scale) were eligible. To calculate the MCID using the distribution-based approach, the standard deviation was estimated using baseline MADRS data while the reliability parameter was measured as the intraclass correlation coefficient between baseline and Week 1. For cross-validation, patients from an observational study were matched to identify the ‘MCID change’ (MADRS change from baseline to endpoint score plus the estimated MCID) and ‘control’ groups. Comparisons of clinical and health-related quality of life measures were performed.Results: In total, 177 placebo-treated patients were identified. MCID estimates for MADRS ranged from 1.6 to 1.9. A total of 105 matched pairs were identified for the cross-validation analyses. Mean change from baseline in MADRS scores (10.6 ± 8.5 vs. 12.5 ± 7.9, p = 0.038) and remission rates (71.6% vs. 57.1%, p < 0.05) significantly differed between the ‘MCID change’ and ‘control’ groups at endpoint. Numerically higher response rates and greater improvements in HRQoL scores in the ‘MCID change’ group were also found.Conclusion: These preliminary findings support the value of the estimated MCID for the MADRS and may aid decision makers in evaluating antidepressant treatment effects and improving long-term patient outcomes.

Estimating species-specific diameter distributions and saw log recoveries of boreal forests from airborne laser scanning data and aerial photographs: a distribution-based approach

The low-density airborne laser scanning (ALS) data based estimation methods have been shown to produce accurate estimates of mean forest characteristics and diameter distributions, according to several studies. The used estimation methods have been based on the laser canopy height distribution approach, where various laser pulse height distribution -derived predictors are related to the stand characteristics of interest. This approach requires very delicate selection methods for selecting the suitable predictor variables. In this study, we introduce a new nearest neighbor search method that requires no complicated selection algorithm for choosing the predictor variables and can be utilized in multipurpose situations. The proposed search method is based on Minkowski distances between the distributions extracted from low density ALS data and aerial photographs. Apart from the introduction of a new search method, the aims of this study were: 1) to produce accurate species-specific diameter distributions and 2) to estimate factual saw log recovery, using the estimated height-diameter distributions and a stem data bank. The results indicate that the proposed method is suitable for producing species-specific diameter distributions and volumes at the stand level. However, it is proposed, that the utilization of more extensive and locally emphasized reference data and auxiliary variables could yield more accurate saw log recoveries.

The minimal important difference of the hospital anxiety and depression scale in patients with chronic obstructive pulmonary disease

BackgroundInterpretation of the Hospital Anxiety and Depression Scale (HADS), commonly used to assess anxiety and depression in COPD patients, is unclear. Since its minimal important difference has never been established, our aim was to determine it using several approaches.Methods88 COPD patients with FEV1 ≤ 50% predicted completed the HADS and other patient-important outcome measures before and after an inpatient respiratory rehabilitation. For the anchor-based approach we determined the correlation between the HADS and the anchors that have an established minimal important difference (Chronic Respiratory Questionnaire [CRQ] and Feeling Thermometer). If correlations were ≥ 0.5 we performed linear regression analyses to predict the minimal important difference from the anchors. As distribution-based approach we used the Effect Size approach.ResultsBased on CRQ emotional function and mastery domain as well as on total scores, the minimal important difference was 1.41 (95% CI 1.18–1.63) and 1.57 (1.37–1.76) for the HADS anxiety score and 1.68 (1.48–1.87) and 1.60 (1.38–1.82) for the HADS total score. Correlations of the HADS depression score and CRQ domain and Feeling Thermometer scores were < 0.5. Based on the Effect Size approach the MID of the HADS anxiety and depression score was 1.32 and 1.40, respectively.ConclusionThe minimal important difference of the HADS is around 1.5 in COPD patients corresponding to a change from baseline of around 20%. It can be used for the planning and interpretation of trials.

Evaluating the reliability, validity and minimally important difference of the Taiwanese version of the diabetes quality of life (DQOL) measurement

BackgroundFew diabetes HRQOL instruments are available in Chinese language. We tested psychometric properties of a Diabetes Quality of Life (DQOL) in Chinese language for diabetes patients in Taiwan and estimated its minimally important differences (MIDs).MethodsData were collected from 337 patients treated in diabetes clinics of a Taiwan teaching hospital. Pearson's correlations among domain scores of the DQOL (satisfaction, impact, and worry), the D-39S (a diabetes-specific instrument, including domains of diabetes control, energy and mobility, social burden and anxiety and worry, and sexual functioning) and the RAND-12 (a generic instrument, including physical health composite (PHC) and mental health composite (MHC)) were estimated to determine convergent/discriminant validity. Known-groups validity was examined using 2-hour postprandial plasma glucose (2 h PPG), hemoglobin A1c (HbA1c)) and presence of complications (retinopathy, neuropathy, and diabetic foot complications rather than the known groups of cardiovascular and cerebrovascular complications). We used a combined anchor- and distribution-based approach to establish MIDs.ResultsThe DQOL scores were more strongly correlated with the physical domains of the D-39S (diabetes control and energy and mobility) and RAND-12 PHC than psychological domains of the D-39S (social burden, anxiety and worry, and sexual functioning) and RAND-12 MHC. The DQOL showed satisfactory discriminative ability for the known groups of 2 h PPG and HbA1c (effect size (ES) ≥ 0.2) and retinopathy, neuropathy, and diabetic foot complications (ES ≥ 0.3), but less satisfactory for the known groups of cardiovascular and cerebrovascular complications. MIDs for the DQOL domains were 3–5 points for satisfaction, 4–5 points for impact, 6–8 points for worry, and 3–4 points for overall HRQOL.ConclusionWe validated a DQOL in Chinese language for diabetes patients in Taiwan and provided MIDs to facilitate the measure of diabetes HRQOL.

Minimal Clinically Important Change of the Neck Disability Index and the Numerical Rating Scale for Patients With Neck Pain

Prospective, single-cohort study. To assess the minimal clinically important change (MCIC) on the Neck Disability Index (NDI) and the Numerical Rating Scale (NRS) for pain in patients with neck pain. Both measurement instruments are frequently used in research and clinical practice, but which changes are clinically relevant is still unknown. The MCIC was estimated with 2 different methods, both integrating an anchor-based and distribution-based approach: the minimal detectable change (MDC) and the optimal cutoff point of the ROC curve. The study population consisted of 183 patients with nonspecific neck pain. The results show an MDC of 10.5 points for the NDI (scale range, 0-50) and 4.3 points for the NRS (scale range, 0-10), and optimal cutoff points of the ROC curve of 3.5 for the NDI and 2.5 for the NRS. The estimated MCIC should be used as an indication for relevant changes in clinical practice. Using the optimal cutoff point of the ROC curve, false positives and false negatives are equally weighted; and if there are no objections doing so, the optimal cutoff point of the ROC curve may be a good choice.

Estimation of minimally important differences in EQ-5D utility and VAS scores in cancer

BackgroundUnderstanding what constitutes an important difference on a HRQL measure is critical to its interpretation. The aim of this study was to provide a range of estimates of minimally important differences (MIDs) in EQ-5D scores in cancer and to determine if estimates are comparable in lung cancer.MethodsA retrospective analysis was conducted on cross-sectional data collected from 534 cancer patients, 50 of whom were lung cancer patients. A range of minimally important differences (MIDs) in EQ-5D index-based utility (UK and US) scores and VAS scores were estimated using both anchor-based and distribution-based (1/2 standard deviation and standard error of the measure) approaches. Groups were anchored using Eastern Cooperative Oncology Group performance status (PS) ratings and FACT-G total score-based quintiles.ResultsFor UK-utility scores, MID estimates based on PS ranged from 0.10 to 0.12 both for all cancers and for lung cancer subgroup. Using FACT-G quintiles, MIDs were 0.09 to 0.10 for all cancers, and 0.07 to 0.08 for lung cancer. For US-utility scores, MIDs ranged from 0.07 to 0.09 grouped by PS for all cancers and for lung cancer; when based on FACT-G quintiles, MIDs were 0.06 to 0.07 in all cancers and 0.05 to 0.06 in lung cancer. MIDs for VAS scores were similar for lung and all cancers, ranging from 8 to 12 (PS) and 7 to 10 (FACT-G quintiles).DiscussionImportant differences in EQ-5D utility and VAS scores were similar for all cancers and lung cancer, with the lower end of the range of estimates closer to the MID, i.e. 0.08 for UK-index scores, 0.06 for US-index scores, and 0.07 for VAS scores.

Methods to Determine the Minimum Important Difference for a Sexual Event Diary Used by Postmenopausal Women with Hypoactive Sexual Desire Disorder

Recently, there has been much discussion in the literature about how to determine the meaningfulness of results generated from a patient-reported outcome measure. A number of reviews have shown that there are two main approaches: anchor- and distribution-based approaches for determining the minimum important difference (MID) for a new measure. There are issues with calculating an MID using each method: Will the two approaches give the same estimate? If the estimates differ, how do you decide on one estimate? Would asking patients directly be more beneficial? A case study was presented to address these issues based on a newly developed diary assessing number of satisfactory sexual events (SSEs) per week in women with hypoactive sexual desire disorder (HSDD). Anchor- and distribution-based estimates were generated from data gathered in two double-blind, placebo-controlled, parallel group trials for the treatment of HSDD (N = 788). A novel interview study was used to ask women directly about an MID for SSEs (N = 77). Defining the MID for an SSE diary in women with HSDD. The estimates varied, producing a range of mean MID estimates between 0.04 and 0.46 SSEs per week. We recommend that rather than defining the MID, a range should be selected from the set of estimates formed by the limits of the 95% confidence intervals.

Intra-Year Shifts in the Earnings Distribution and Implications for Earnings Management*

We extend the earnings distribution-based approach to explain the kink in the annual earnings distribution by examining its evolution. Our analyses fail to confirm prior research implications that the kink develops when firms facing smallest year-to-date losses abnormally shift into the smallest profit interval. We find the primary abnormal tendency by firms during the development of the kink is to remain in the smallest profit interval, rather than to move into a higher earnings interval. This suggests that firms avoiding smallest losses most likely prevent year-to-date profits from turning into annual losses rather than converting year-to-date losses into annual profits.

Responsiveness and minimally important change for the Manchester-Oxford foot questionnaire (MOXFQ) compared with AOFAS and SF-36 assessments following surgery for hallux valgus

To assess responsiveness and minimally important change (MIC) for the Manchester-Oxford foot questionnaire (MOXFQ) using anchor and distribution-based approaches. Responsiveness and estimates of minimal clinically important difference (MCID) and minimal detectable change are compared with those from the Short-Form 36 (SF-36) and American Orthopaedic Foot & Ankle Society (AOFAS) measures. A prospective observational study of 91 consecutive patients (125 foot operations) undergoing hallux valgus surgery at an orthopaedic hospital. Pre- and 12 month post-surgery, patients completed the MOXFQ and SF-36, and foot surgeons assessed all four AOFAS scores corresponding to four regions of the foot. Transition items were asked about perceived changes compared with before surgery. Mean changes in all domains of each instrument were statistically significant, but foot-specific MOXFQ and AOFAS domains produced much larger effect sizes (>1) than any SF-36 domains, indicating superior responsiveness. Clear associations occurred between transition items and all MOXFQ and AOFAS scores, but with only one (physical function) SF-36 domain. Anchor and distribution-based approaches identified generally comparable measures of MIC, which for the MOXFQ and AOFAS domains were between 1 and 2 standard error of measurement. In metric terms, the MCIDs were 16, 12, and 24 for the MOXFQ Walking/standing, Pain, and Social Interaction domains, respectively. For hallux valgus surgery, the MOXFQ is highly responsive. Performance is comparable to the AOFAS and notably better than the generic SF-36. Study estimates of MIC for the MOXFQ are useful to inform sample-size calculations for future clinical trials.

Minimally important change determined by a visual method integrating an anchor-based and a distribution-based approach

Background:Minimally important changes (MIC) in scores help interpret results from health status instruments. Various distribution-based and anchor-based approaches have been proposed to assess MIC.Objectives:To describe and apply a visual method, called the anchor-based MIC distribution method, which integrates both approaches.Method:Using an anchor, patients are categorized as persons with an important improvement, an important deterioration, or without important change. For these three groups the distribution of the change scores on the health status instrument are depicted in a graph. We present two cut-off points for an MIC: the ROC cut-off point and the 95% limit cut-off point.Results:We illustrate our anchor-based MIC distribution method determining the MIC for the Pain Intensity Numerical Rating Scale in patients with low back pain, using two conceivable definitions of minimal important change on the anchor. The graph shows the distribution of the scores of the health status instrument for the relevant categories on the anchor, and also the consequences of choosing the ROC cut-off point or the 95% limit cut-off point.Discussion:The anchor-based MIC distribution method provides a general framework, applicable to all kind of anchors. This method forces researchers to choose and justify their choice of an appropriate anchor and to define minimal importance on that anchor. The MIC is not an invariable characteristic of a measurement instrument, but may depend, among other things, on the perspective from which minimal importance is considered and the baseline values on the measurement instrument under study. A balance needs to be struck between the practicality of a single MIC value and the validity of a range of MIC values.

Effects of pesticides on soil invertebrates in laboratory studies: A review and analysis using species sensitivity distributions

Species sensitivity distributions (SSD) and 5% hazardous concentrations (HC5) are distribution-based approaches for assessing environmental risks of pollutants. These methods have potential for application in pesticide risk assessments, but their applicability for assessing pesticide risks to soil invertebrate communities has not been evaluated. Using data obtained in a systematic review, the present study investigates the relevance of SSD and HC5 for predicting pesticide risks to soil invertebrates. Altogether, 1950 laboratory toxicity data were obtained, representing 250 pesticides and 67 invertebrate taxa. The majority (96%) of pesticides have toxicity data for fewer than five species. Based on a minimum of five species, the best available endpoint data (acute mortality median lethal concentration) enabled SSD and HC5 to be calculated for 11 pesticides (atrazine, carbendazim, chlorpyrifos, copper compounds, diazinon, dimethoate, gamma-hexachlorocyclohexane, lambda-cyhalothrin, parathion, pentachlorophenol, and propoxur). Arthropods and oligochaetes exhibit pronounced differences in their sensitivity to most of these pesticides. The standard test earthworm species, Eisenia fetida sensu lato, is the species that is least sensitive to insecticides based on acute mortality, whereas the standard Collembola test species, Folsomia candida, is among the most sensitive species for a broad range of toxic modes of action (biocide, fungicide, herbicide, and insecticide). These findings suggest that soil arthropods should be tested routinely in regulatory risk assessments. In addition, the data indicate that the uncertainty factor for earthworm acute mortality tests (i.e., 10) does not fully cover the range of earthworm species sensitivities and that acute mortality tests would not provide the most sensitive risk estimate for earthworms in the majority (95%) of cases.

Use of Interval Arithmetic to Incorporate the Uncertainty of Load Demand for Radial Distribution System Analysis

In this letter, the probabilistic distribution-based interval arithmetic approach is used to incorporate uncertainty in load demand. The variation of load is represented as Gaussian distribution function. It is suggested that the proposed power flow could be useful for planning and expansion planning of radial distribution systems.

Distribution-based and anchor-based approaches provided different interpretability estimates for the Hydrocephalus Outcome Questionnaire

Objective To compare three separate methods for establishing interpretability for a health status measure, the Hydrocephalus Outcome Questionnaire (HOQ). Study Design and Setting The mothers of children with hydrocephalus attending the outpatient clinics at a pediatric hospital completed the HOQ (for which scores can range from 0 to 1.0 with the smallest possible incremental change being .005), the Health Utilities Index-2 (HUI-2), and a global rating of their child's health. The surgeon for the child also provided a global rating of the child's health following their visit. These data were used to calculate (i) the minimal important difference (MID) based on global health ratings, (ii) the MID based on an effect size approach, and (iii) the conversion of numerical HOQ scores into health utility scores obtained from the HUI-2. Results Based on mothers' responses ( n = 79) and surgeons' responses ( n = 61), respectively, the MID for the HOQ was estimated to be .12 and .10. Using the effect size approach, the MID was estimated to be much lower at .03. HOQ scores were found to be readily translatable to HUI-2 utility scores using a simple linear transformation. The mean utility score for this sample of patients was .77. Conclusions Two methods for determining the MID yielded quantitatively different results. Conversion of numerical health status scores to utility scores was done successfully and providing another element of interpretability.

PROBABILISTIC DISTRIBUTION BASED INTERVAL ARITHMETIC LOAD FLOW ALGORITHM FOR RADIAL DISTRIBUTION SYSTEM

Load flow is an essential tool for the study of radial distribution systems (RDS). The inputs to load flow solutions are subject to uncertainties due to load and line parameter variations. Unfortunately, load flow research has to date considered only fixed load and line parameters; inherent uncertainties of the inputs were overlooked. In this article probabilistic distribution-based interval arithmetic approach is proposed to incorporate variations in load parameter. The uncertainties of load variation are represented as Gaussian distribution function. The proposed load flow is tested on standard test systems; the authors find that solutions obtained provide much wider information and all possible solution states are obtained in the closed bounded interval form. It is suggested that proposed power flow could be useful for planning and expansion planning of RDS, where future data always carry a high degree of uncertainties. Composite load model is also incorporated in the algorithm.

SUBSPACE IDENTIFICATION OF THE SYSID2006 BENCHMARKS VIA DISTRIBUTION-BASED APPROACH

This paper presents solutions to the IFAC SYSID2006 Benchmark identification problem by applying the subspace identification via distribution-based approach proposed in Ohsumi, Kameyama and Yamaguchi (2002). The application of the algorithm seems to have succeeded in establishing continuous-time models for the benchmark data.

The neurologist in the DGH: how to survive without on site neuroradiolgy

<h3>Background</h3> The Epworth Sleepiness Scale (ESS) is a widely used tool for assessing sleepiness in patients with obstructive sleep apnoea (OSA). We aimed to estimate the minimal important difference (MID)...

Fuzzy Set Inclusion

The concept of set inclusion has remained insufficiently developed in the fuzzy set literature to be of much use to social scientists. However, a fully fledged concept of fuzzy set inclusion, along with appropriate statistical methods for evaluating it, could be very useful in the social sciences. This article combines fuzzy set and statistical methods, in the form of a cumulative distribution-based approach to evaluating fuzzy set inclusion without making strong assumptions about measurement levels. It establishes criteria for distinguishing an “inclusion relation” from independence plus skew as well as from other kinds of relationships. A measure of inclusion is developed that is sensitive to the degree to which individual cases violate a strict inclusion rule. A technique for modeling localized inclusion relations in contingency tables and scatter plots is also presented. Finally, the connections between the fuzzy set approach to set inclusion and mainstream statistical techniques are briefly adumbrated.

Development and Validation of a Patient Satisfaction and Preference Questionnaire for Inhalation Devices

The Patient Satisfaction and Preference Questionnaire (PASAPQ) is a multi-item measure of respiratory inhalation device satisfaction and preference designed to be easily understood and administered to patients with asthma and COPD. This study assessed its validity, reliability and responsiveness and explored the between-group difference in PASAPQ scores that is meaningful. The field test version was developed using literature, focus groups and expert opinion. Item reduction followed. The assessment of the validity, reliability and responsiveness of the PASAPQ utilized data from two clinical studies comparing devices delivering the same medication, and was performed with pre-specified criteria. A minimally important difference (MID) was estimated using both anchor- and distribution-based approaches. Two factors of the PASAPQ, 'performance' and 'convenience', were consistent across studies. Missing and out-of-range data were minimal (<1%) and respondents used a full range of response options. All items correlated most highly with their hypothesized scale and all exceeded the minimum correlation criteria of 0.40. Cronbach's alfa was high (0.87-0.94), providing support for internal reliability for the PASAPQ. Correlations of the overall satisfaction item with the performance domain ranged from 0.78 to 0.91, the convenience domain ranged from 0.54 to 0.71, and the total score ranged from 0.78 to 0.90. These moderate-to-strong correlations provide substantial support for the validity of the PASAPQ domains and total score. Discriminate validity was assessed by calculating PASAPQ scores for patients' ratings of the device that they preferred compared with the other, non-preferred device. The preferred device was rated higher on all satisfaction measures, supporting the ability of the PASAPQ to discriminate between preferred and non-preferred devices. Although a difference of 3 or 4 points may be sufficient to observe a small effect difference between groups, most of the MID estimates were in the 8-10 point range. Our analyses across asthma, COPD and patients with mixed respiratory disease (with features of both COPD and asthma), study designs and data sets lead us to conclude that the PASAPQ is a practical, valid, reliable and responsive instrument for measuring respiratory device satisfaction. Furthermore, a difference in satisfaction scores between treatment groups of 10 points is, conservatively, a difference that is meaningful to patients.

Minimally Clinically Important Difference for the UCSD Shortness of Breath Questionnaire, Borg Scale, and Visual Analog Scale

Dyspnea is a primary symptom of chronic lung disease and an important outcome measure for clinical trials. Several standardized measures have been developed to evaluate this important symptom and are being used increasingly in clinical trials. The minimally clinically important difference (MCID) is not well defined for these measures but is important in interpreting the clinical meaning of results of studies in this area. The purpose of this paper is to evaluate the MCID for three commonly used measures to assess dyspnea in chronic lung disease: UCSD Shortness of Breath Questionnaire (SOBQ), Borg Scale (Borg), and Visual Analog Scale (VAS). The analysis is based on a retrospective review of published trials evaluating the response to a pulmonary rehabilitation or exercise intervention that is known to produce modest, but clinically meaningful changes for such patients. Using a distribution-based approach based primarily on effect size, the recommended MCID for these measures are: 5-units for the SOBQ, 1-unit for the Borg scale, and approximately 10 to 20 units for the VAS.

How much is enough and who says so?

One of the challenges of health-related quality of life research is to translate statistically significant health-related quality of life changes into interpretable clinical or medically important ones. To calculate the minimal important difference of the King's Health Questionnaire, a condition-specific health-related quality of life questionnaire for the assessment of men and women with lower urinary tract dysfunction. The King's Health Questionnaire was administered to patients suffering from overactive bladder enrolled in two multinational studies. Minimal important differences were calculated using an anchor-based approach with both a global rating of patient-perceived treatment benefit and one of perceived disease impact. A distribution-based method using effect size was calculated for comparison purposes. Minimal important difference values varied slightly with each method. Using the anchor-based approach, the King's Health Questionnaire minimal important difference ranged between 5-10 points when the calculation factored out patients who reported no change and 6-12 points for patients who experienced a small improvement. The effect size method indicated a minimal important difference of 5 to 6 points for a small effect and 10 to 15 points for a medium effect. In the case of the King's Health Questionnaire, the anchor-based approaches and the distribution-based approach provide similar results. A change from baseline of at least 5 points on King's Health Questionnaire domains indicates a change that is meaningful to patients and is indicative of a clinically meaningful improvement in health-related quality of life after treatment. Convergence of the estimates using different approaches should give us confidence in the values derived for the quality of life domains measured by the King's Health Questionnaire.