Predicting missing quality of life data that were later recovered: an empirical comparison of approaches

Background and Purpose The aim was to compare simple imputation, multiple imputation, and modeling approaches to deal with ‘missing’ quality of life data. Data were obtained from five clinical trials, which employed a reminder system for follow-up questionnaires. Previous studies have compared imputation strategies by artificially removing data according to prespecified mechanisms. Our approach differs from previous study as actual collected data are utilized.

Methods Data obtained by reminder were initially treated as missing. These missing values were imputed using a variety of simple and multiple imputation strategies. The trials were analyzed using the imputed datasets, and the resulting treatment effects compared to analyses using the full dataset including responses following reminders. A repeated measures model was also carried out on the available data and the pattern mixture models were employed. The accuracy of the different strategies was assessed by calculating the bias seen in the calculated treatment difference compared to the actual observed treatment difference.

Results Baseline carried forward or last value carried forward were shown to be the best simple imputation methods in this setting. Multiple imputation using a regression model or predictive mean match model tended to provide treatment difference estimates with the least bias when compared to the actual observed data. Pattern mixture models did not perform well. Overall, the multiple imputation procedures were generally the least biased approaches.

Limitations A number of imputation and modeling procedures have been investigated but this list is not exhaustive. All the example datasets come from the same data source and perhaps studies from additional disease areas would have been useful. However, we feel the results are generalizable to other quality of life outcomes and clinical areas.

Conclusions Multiple imputation is recommended for missing quality of life data as it makes the assumption of missing at random which in the quality of life setting is more plausible than the assumption of missing completely at random for which most simple imputation methods are based. Pattern mixture models can be complex and did not perform well in this setting. Clinical Trials 2010; 7: 333—342. http://ctj.sagepub.com

Development of computerised adaptive testing (CAT) for the EORTC QLQ-C30 dimensions - general approach and initial results for physical functioning

BACKGROUND

Health-related quality of life (HRQOL) questionnaires should ideally be adapted to the individual patient and at the same time scores should be directly comparable across patients. This is achievable using a computerised adaptive test (CAT). Basing the CAT on an existing instrument enables measurement within an established HRQOL framework and allows backward-compatibility with studies using the original instrument. Because of these advantages the EORTC Quality of Life Group (QLG) has initiated a project to develop a CAT version of the widely used EORTC QLQ-C30.

METHODS

We present the EORTC QLG's strategy for developing a CAT. For each dimension of the EORTC QLQ-C30 our approach includes literature search and conceptualisation, formulation of new items, expert and patient evaluations, field-testing, and psychometric analyses of the items. The strategy is illustrated with the initial results of the development of CAT for physical functioning (PF).

RESULTS

We identified 975 PF items in the literature. Of these, 407 items were deemed relevant, i.e. measured one of the PF aspects measured by the QLQ-C30. Based on these items we developed 86 new items. Review by the EORTC CAT-project group reduced this to 66 items. Based on expert and patient evaluations several items were revised and the list was further reduced to 51 items.

CONCLUSIONS

Based on the findings for PF, we believe that our approach will generate item pools that are relevant and appropriate for cancer patients. These will form the basis for a backward-compatible CAT assessing the HRQOL dimensions of the EORTC QLQ-C30.

Assessment of body image in patients undergoing surgery for colorectal cancer

PURPOSE

This study tested the scale properties and validity of the ten-item body image scale (BIS) in patients undergoing surgery for colorectal cancer (CRC).

METHODS

Patients completed the BIS and a validated measure of health-related quality of life (European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30) after open or laparoscopic resection. A sample of the patients had also previously completed questionnaires before and after surgery. Multi-trait scaling and factor analysis were used to examine the questionnaire scaling, and tests of reliability, clinical and construct validity were performed.

RESULTS

Eighty-two patients (48 male, 59 open, 23 laparoscopic) participated. Scaling analyses suggested a nine-item scale and one single item. This revised scale structure demonstrated good test-retest reliability (r = 0.94) and no overlap with the key domains of the EORTC QLQ-C30 (r < 0.40). Patients with a stoma reported significantly poorer BIS scores than those undergoing simple resection (p = 0.005).

CONCLUSIONS

This study provides psychometric and clinical evidence for a revised scale structure for the BIS in CRC, and the questionnaire is suitable to assess body image in CRC clinical trials.

Validation of Doloplus-2 among nonverbal nursing home patients--an evaluation of Doloplus-2 in a clinical setting

BACKGROUND

Pain measurement in nonverbal older adults is best based on behavioural observation, e.g. using an observational measurement tool such as Doloplus-2. The purposes of this study were to examine the use of Doloplus-2 in a nonverbal nursing home population, and to evaluate its reliability and validity by comparing registered nurses' estimation of pain with Doloplus-2 scores.

METHOD

In this cross-sectional study, Doloplus-2 was used to observe the pain behaviour of patients aged above 65 years who were unable to self-report their pain. Nurses also recorded their perceptions of patient pain (yes, no, don't know) before they used Doloplus-2. Data on demographics, medical diagnoses, and prescribed pain treatment were collected from patient records. Daily life functioning was measured and participants were screened using the Mini Mental State Examination.

RESULTS

In total, 77 nursing home patients were included, 75% were women and the mean age was 86 years (SD 6.6, range 68-100). Over 50% were dependent on nursing care to a high or a medium degree, and all were severely cognitively impaired. The percentage of zero scores on Doloplus-2 ranged from 17% (somatic reactions) to 40% (psychosocial reactions). Cronbach's alpha was 0.71 for the total scale. In total, 52% of the patients were judged by nurses to be experiencing pain, compared with 68% when using Doloplus-2 (p = 0.01). For 29% of the sample, nurses were unable to report if the patients were in pain.

CONCLUSIONS

In the present study, more patients were categorized as having pain while using Doloplus-2 compared with nurses' estimation of pain without using any tools. The fact that nurses could not report if the patients were in pain in one third of the patients supports the claim that Doloplus-2 is a useful supplement for estimating pain in this population. However, nurses must use their clinical experience in addition to the use of Doloplus-2, as behaviour can have different meaning for different patients. Further research is still needed about the use of Doloplus-2 in patients not able to self-report their pain.

A systematic review of oxymorphone in the management of chronic pain

Opioids are recommended for control of moderate-to-severe, chronic, malignant, and nonmalignant pain. A controlled-release formulation of the opioid oxymorphone has recently been launched. The aim of this review was to assess the effectiveness of oxymorphone as an analgesic in chronic pain. A systematic search for published studies of oral oxymorphone in the management of chronic pain was conducted. The studies were evaluated for their internal validity according to standard criteria. They were also evaluated for their external validity and research ethic aspects. A meta-analysis was performed to examine the effect of oxymorphone compared with placebo. Nine studies were evaluated; three were excluded because of low quality. Six controlled studies (duration 2-12 weeks) included a total of 1489 subjects suffering from chronic low back pain, chronic pain from osteoarthritis, and chronic cancer pain. Three of the studies were of high quality and three of medium quality. External validity was assessed to be high, medium, and low (in one, three, and two studies, respectively). The meta-analysis suggests that daily doses of 40-100mg are superior to placebo; however, the estimate (reduction of pain intensity compared with placebo) of the treatment effect is imprecise (95% confidence interval -17.08, -8.69). Limited evidence suggests that oxymorphone is effective for pain control in patients with cancer. No significant differences between oxymorphone and oxycodone at equipotent doses were found. In conclusion, oxymorphone is superior to placebo. There is no evidence that the efficacy of oxymorphone differs from other opioids.

What changes in health-related quality of life matter to multiple myeloma patients? A prospective study

OBJECTIVE

To determine the clinical significance of changes in quality-of-life scores in patients with multiple myeloma (MM), we have estimated the minimal important difference (MID) for the health-related quality-of-life instrument, the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30. The MID is the smallest change in a quality-of-life score considered important to patients.

METHODS

Between 2006 and 2008, 239 patients with MM completed the EORTC QLQ-C30 at inclusion (T1) and after 3 months (T2). At T2, a structured quality-of-life interview was also performed. MIDs were calculated by using mean score changes (T2-T1) for patients who in the interview stated they had improved, deteriorated or were unchanged. MIDs were also estimated by the receiver-operating characteristic (ROC) curve method as well as by calculation effect sizes using standard deviations of baseline scores.

RESULTS

MIDs varied slightly depending on the method used. Patients stating in the interview that they had 'improved' or 'deteriorated' had a corresponding change in EORTC QLQ-C30 score ranging from 6 to15 (improvement) and from 9 to17 (deterioration) (scale range 0-100). The ROC analysis indicated that changes in score from 7 to17 represent clinically important changes to patients. The effect size method suggested 5-6 to be a small and 11-15 to be a medium change.

CONCLUSION

Calculation of MIDs as mean score changes or by ROC analysis suggested that a change in the EORTC QLQ-C30 score in the range of approximately 6-17 is considered important by patients with MM. These MIDs are closer to a medium effect size than to a small effect size. Our findings imply that mean score changes smaller than 6 are unlikely to be important to the patients, even if these changes are statistically significant.

Health-related quality of life assessment in randomised controlled trials in multiple myeloma: a critical review of methodology and impact on treatment recommendations

OBJECTIVES

Patients with multiple myeloma (MM) often have pronounced symptoms and substantially reduced quality of life. The aims of treatment are to control disease, maximise quality of life and prolong survival. Hence, health-related quality of life (HRQOL) should be an important end-point in randomised controlled trials (RCTs) in addition to traditional endpoints. We wanted to evaluate whether trials reporting HRQOL outcomes have influenced clinical decision making and whether HRQOL was assessed robustly according to predefined criteria.

METHODS

A systematic review identified RCTs in MM with HRQOL assessment as a study end-point. The methodological quality of these studies was assessed according to a checklist developed for evaluating HRQOL outcomes in clinical trials. The impact of the HRQOL results on clinical decision making was assessed, using published clinical guidelines as a reference.

RESULTS

Fifteen publications presenting RCTs with HRQOL as a study end-point were identified. In 13 trials, the author stated that HRQOL results should influence clinical decision making. We found, however, that the HRQOL data only had a limited impact on published treatment guidelines for bisphosphonates, high-dose treatment, interferon, erythropoiesis-stimulating agents and novel agents.

CONCLUSION

The present review indicates that the there are still few RCTs in MM including HRQOL as a study end-point. Systematic incorporation of HRQOL measures into clinical trials allows for a comparison of treatment arms that includes the patients' perspective. Before the full impact on clinical decisions can be realised, the quality and methodology of collecting HRQOL data must be further improved and the results rendered more comprehensible to clinicians.

Validation of the European Organization for Research and Treatment of Cancer QLQ-LMC21 questionnaire for assessment of patient-reported outcomes during treatment of colorectal liver metastases

Background

There is a lack of valid patient-reported outcome (PRO) measures for hepatectomy or palliative treatment of colorectal hepatic metastases. This study examined the validity and reliability of the European Organization for Research and Treatment of Cancer Quality of Life questionnaire liver module (QLQ-LMC21) in patients with liver metastases from colorectal cancer.

Methods

Some 356 patients completed the core questionnaire (QLQ-C30), QLQ-LMC21 and a debriefing questionnaire before and 3 months after hepatectomy or palliative treatment. Construct, criterion and clinical validity were evaluated before and after treatment.

Results

Questionnaire compliance was high; less than 1 per cent of data were missing from individual items. Modifications to the hypothesized scale structure produced four scales and nine single items with good reliability, clinical, criterion and construct validity. The QLQ-LMC21 distinguished between patients selected for surgery or palliative treatment in nine of these 13. Significant changes in PROs were observed before and after treatment. The new module discriminated between clinically distinct groups of patients and measured aspects of quality of life not covered in the core questionnaire.

Conclusion

The EORTC QLQ-LMC21 is a valid and reliable questionnaire module to use with the QLQ-C30 in assessing PROs in hepatectomy or palliative treatment for colorectal liver metastases.

Symptom Assessment in Palliative Care: A Need for International Collaboration

This article describes the research strategy for the development of a computerized assessment tool as part of a European Union (EU)–funded project, the European Palliative Care Research Collaborative (EPCRC). The EPCRC is funded through the Sixth Framework Program of the EU with major objectives to develop a computer-based assessment and classification tool for pain, depression, and cachexia. A systematic approach will be applied for the tool development with emphasis on multicultural and multilanguage challenges across Europe. The EPCRC is based on a long lasting collaboration within the European Association for Palliative Care Research Network. The ongoing change in society towards greatly increased use of communication as well as information transfer via digital systems will rapidly change the health care system. Therefore, patient-centered outcome assessment tools applicable for both clinic and research should be developed. Report of symptoms via digital media provides a start for face-to-face communication, treatment decisions, and assessment of treatment effects. The increased use of electronic media for exchange of information may facilitate the development and use of electronic assessment tools and decision-making systems in oncology. In the future, patients may find that a combination of a face-to-face interview plus a transfer of information of subjective symptoms by electronic means will optimize treatment.

Chronic non-malignant pain patients report as poor health-related quality of life as palliative cancer patients

BACKGROUND

Patients with chronic non-malignant pain (CNMP) conditions are known to report reduced health-related quality of life (HRQoL). The objective of this exploratory study was to compare HRQoL between patients admitted to a multidisciplinary pain centre, palliative cancer (PC) patients and national norms.

METHODS

HRQoL data from 288 patients with CNMP admitted to the multidisciplinary pain centre at Trondheim University Hospital were compared with 434 patients with advanced cancer included in a trial of comprehensive palliative care in the hospital palliative medicine unit and national norms. HRQoL was assessed using the EORTC QLQ-C30. Age- and gender-adjusted norm data were calculated and compared between the two groups.

RESULTS

Scores from both groups deviated from adjusted norm data on all scales, with poorer functioning and more symptoms. Compared with PC patients, CNMP patients reported a larger deviation (worse scores) on global quality of life, cognitive functioning, pain, sleep disturbances and financial difficulties. Deviations from norm data were similar for physical, social and emotional functioning, diarrhoea, dyspnoea and fatigue. PC patients reported worse scores on role functioning, nausea/vomiting, loss of appetite and constipation.

CONCLUSION

CNMP patients admitted to multidisciplinary pain centres report significantly reduced HRQoL, in addition to severe pain. They consider their HRQoL to be as poor as HRQoL reported from dying cancer patients and substantially poorer than national norms. Factors other than the biological severity of the disease seem to be of major importance for self-reported HRQoL.

Clinical and psychometric validation of a questionnaire module, the EORTC QLQ-OG25, to assess health-related quality of life in patients with cancer of the oesophagus, the oesophago–gastric junction and the stomach

AIM

To combine and test the EORTC questionnaires for assessing quality of life (HRQL) for oesophageal (QLQ-OES18) and stomach cancer (QLQ-STO22), into a single questionnaire for tumours of the oesophagus, oesophago-gastric junction or stomach.

METHODS

The QLQ-OES18, QLQ-STO22 and seven modified items were administered to 300 patients with oesophageal (n=148), junctional (n=66), or gastric cancer (n=86). Semi-structured interviews assessed item and scale preference and multi-trait scaling analyses confirmed the scale structure of the new module (QLQ-OG25). This was further tested for validity.

RESULTS

The QLQ-OG25 has six scales, dysphagia, eating restrictions, reflux, odynophagia, pain and anxiety. Scales have good reliability (alpha range 0.67-0.87) and they distinguish between tumour sites and disease stage. Scales do not correlate highly with scores from the core questionnaire, thus indicating that the module was addressing separate HRQL aspects.

CONCLUSION

The QLQ-OG25 is recommended to supplement the EORTC QLQ-C30 when assessing HRQL in patients with oesophageal, junctional or gastric cancer.

Translation procedures for standardised quality of life questionnaires: The European Organisation for Research and Treatment of Cancer (EORTC) approach

BACKGROUND

The European Organisation for Research and Treatment of Cancer quality of life (EORTC QL) questionnaires are used in international trials and therefore standardised translation procedures are required. This report summarises the EORTC translation procedure, recent accomplishments and challenges.

METHODS

Translations follow a forward-backward procedure, independently carried out by two native-speakers of the target language. Discrepancies are arbitrated by a third consultant, and solutions are reached by consensus. Translated questionnaires undergo a pilot-testing. Suggestions are incorporated into the final questionnaire. Requests for translations originate from the module developers, physicians or pharmaceutical industry, and most translations are performed by professional translators. The translation procedure is managed and supervised by a Translation Coordinator within the EORTC QL Unit in Brussels.

RESULTS

To date, the EORTC QLQ-C30 has been translated and validated into more than 60 languages, with further translations in progress. Translations include all major Western, and many African and Asian languages. The following translation problems were encountered: lack of expressions for specific symptoms in various languages, the use of old-fashioned language, recent spelling reforms in several European countries and different priorities of social issues between Western and Eastern cultures. The EORTC measurement system is now registered for use in over 9000 clinical trials worldwide.

CONCLUSIONS

The EORTC provides strong infrastructure and quality control to produce robust translated questionnaires. Nevertheless, translation problems have been identified. The key to improvements may lie in the particular features and strengths of the group, consisting of researchers from 21 countries representing 25 languages and include the development of simple source versions, the use of advanced computerised tools, rigorous pilot-testing, certification procedures and insights from a unique cross-cultural database of nearly 40,000 questionnaire responses.

Using the literature to quantify the learning curve: A case study

Objectives: The aim of this study was to assess whether a literature review of a technology can allow a learning curve to be quantified.

Methods: The literature for fiberoptic intubation was searched for studies reporting information relevant to the learning curve. The Cochrane Library, Medline, Embase, and Science Citation index were searched. Studies that reported a procedure time were included. Data were abstracted on the three features of learning: initial level, rate of learning, and asymptote level. Random effects meta-analysis was performed.

Results: Only twenty-one studies gave explicit information concerning the previous experience of the operator(s). There were thirty-two different definitions of procedure time. From four studies of fiberoptic nasotracheal intubation, the mean starting level and time for the 10th procedure was estimated to be 133 seconds (95 percent confidence interval, 113–153) and 71 seconds (95 percent confidence interval, 62–79), respectively.

Conclusions: The review approach allowed learning to be quantified for our example technology. Poor and insufficient reporting constrained formal statistical estimation. Standardized reporting of nondrug techniques with adequate learning curve details is needed to inform trial design and cost-effectiveness analysis.

Randomised trial of in-hospital geriatric intervention: impact on function and morale

BACKGROUND

In two previous publications, we have shown that treatment of acutely sick, frail elderly patients in a Geriatric Evaluation and Management Unit (GEMU) compared to treatment in the general Medical Wards (MW) reduced mortality and improved the chances of living at home in contrast to living in nursing homes or being dead.

OBJECTIVE

The aim of this presentation was to study the impact on function, symptoms of depression and general well-being of treatment in the GEMU as compared to treatment in MW.

METHODS

Acutely sick, frail patients aged >or=75 years, admitted as emergencies to the Department of Internal Medicine, were randomised either to treatment in the GEMU (n = 127) or the MW (n = 127). In the GEMU the treatment strategy emphasised comprehensive interdisciplinary assessment of all relevant disorders, prevention of complications and iatrogenic conditions, early mobilisation, rehabilitation and discharge planning. The control group received treatment as usual from the Department of Internal Medicine. After discharge neither group received specific follow-up. Activities of daily living (ADL), instrumental ADL, cognitive function, symptoms of depression and general well-being were assessed 3, 6 and 12 months after discharge from hospital.

RESULTS

There was no difference in function, depression or general well-being in the GEMU as compared to the MW group. If the dead were included in the analysis at the highest ADL dependency level, there was better function in the GEMU group at 3 months (p = 0.03).

CONCLUSION

Treatment in the GEMU had no measurable beneficial impact on function, morale or symptoms of depression. Taken the previously shown mortality reduction into consideration an additional effect on function was less likely and the overall treatment effect was considered to be positive.

Health-related quality of life measurement in randomized clinical trials in surgical oncology

PURPOSE

There is debate about the value of measuring health-related quality of life (HRQL) in clinical trials in oncology because of evidence suggesting that HRQL does not influence clinical decisions. Analysis of HRQL in surgical trials, however, may inform decision making because it provides detailed assessment of the immediate detrimental short-term impact of surgery on HRQL that needs to be considered against the long-term survival benefits and functional outcomes of surgery. This study evaluated whether HRQL in randomized trials in surgical oncology contributes to clinical decision making.

METHODS

A systematic review identified randomized trials in surgical oncology with HRQL. Trials were evaluated independently by two reviewers and the value of HRQL in clinical decision making was categorized in three ways: whether trial investigators reported that HRQL influenced final treatment recommendations, whether trial investigators reported that HRQL would be useful for informed consent, and whether HRQL was assessed robustly according to predefined criteria.

RESULTS

Thirty-three randomized trials with valid HRQL questionnaires were identified; 22 (67%) concluded that HRQL outcomes influenced treatment decisions or provided valuable data for informed consent, and seven of these trials had robust HRQL design. Another five trials had robust HRQL design but investigators reported that HRQL outcomes were not clinically important enough to influence treatment recommendations.

CONCLUSION

In surgical trials in oncology, HRQL informed clinical decision making. It is recommended that HRQL be included in relevant surgical trials, and that information be used to inform clinicians and patients about the impact of surgery on short- and long-term HRQL.

Prospective randomised multicenter trial on single fraction radiotherapy (8Gy×1) versus multiple fractions (3Gy×10) in the treatment of painful bone metastases

BACKGROUND AND PURPOSE

To investigate whether single-fraction radiotherapy is equal to multiple fractions in the treatment of painful metastases.

PATIENTS AND METHODS

The study planned to recruit 1000 patients with painful bone metastases from four Norwegian and six Swedish hospitals. Patients were randomized to single-fraction (8 Gy x 1) or multiple-fraction (3 Gy x 10) radiotherapy. The primary endpoint of the study was pain relief, with fatigue and global quality of life as the secondary endpoints.

RESULTS

The data monitoring committee recommended closure of the study after 376 patients had been recruited because interim analyses indicated that, as in two other recently published trials, the treatment groups had similar outcomes. Both groups experienced similar pain relief within the first 4 months, and this was maintained throughout the 28-week follow-up. No differences were found for fatigue and global quality of life. Survival was similar in both groups, with median survival of 8-9 months.

CONCLUSIONS

Single-fraction 8 Gy and multiple-fraction radiotherapy provide similar pain benefit. These results, confirming those of other studies, indicate that single-fraction 8 Gy should be standard management policy for these patients.

Multidimensional Computerized Adaptive Testing of the EORTC QLQ-C30: Basic Developments and Evaluations

OBJECTIVE

Self-report questionnaires are widely used to measure health-related quality of life (HRQOL). Ideally, such questionnaires should be adapted to the individual patient and at the same time scores should be directly comparable across patients. This may be achieved using computerized adaptive testing (CAT). Usually, CAT is carried out for a single domain at a time. However, many HRQOL domains are highly correlated. Multidimensional CAT may utilize these correlations to improve measurement efficiency. We investigated the possible advantages and difficulties of multidimensional CAT.

STUDY DESIGN AND SETTING

We evaluated multidimensional CAT of three scales from the EORTC QLQ-C30: the physical functioning, emotional functioning, and fatigue scales. Analyses utilised a database with 2958 European cancer patients.

RESULTS

It was possible to obtain scores for the three domains with five to seven items administered using multidimensional CAT that were very close to the scores obtained using all 12 items and with no or little loss of measurement precision.

CONCLUSION

The findings suggest that multidimensional CAT may significantly improve measurement precision and efficiency and encourage further research into multidimensional CAT. Particularly, the estimation of the model underlying the multidimensional CAT and the conceptual aspects need further investigations.

The relationship between patient knowledge of hemoglobin levels and health-related quality of life

BACKGROUND

An anti-anemia drug may improve self-reported quality of life (QOL) partly because patients know their hemoglobin level is rising. In the absence of any published studies on this topic, the authors investigated the association between knowledge of hemoglobin levels and self-reported QOL.

METHODS

The study analyzed health-related QOL (HRQOL) data from five randomized clinical trials of erythropoietic therapy in patients with cancer-related anemia. Patients were asked whether they knew their hemoglobin level and, if so, to report its value. Patients (n=1007) were grouped into three categories depending on the extent and accuracy of hemoglobin level knowledge. HRQOL scale scores were compared between categories.

RESULTS

Only 23.2% of patients reported knowing their hemoglobin level at the end of the study; however, the value was accurate (within 1 g/dl) in 88.0% of these patients. On five of the 11 HRQOL scales studied, there was a significant association between knowledge of hemoglobin level and HRQOL score. However, the magnitude of the mean difference between those who knew vs. those who did not know their hemoglobin was generally below scale thresholds for minimally important differences.

CONCLUSIONS

Patient knowledge of hemoglobin level has a modest association with some aspects of self-reported HRQOL. The magnitude of this association, where it exists, would be unlikely to explain large group differences in HRQOL reports over time, even for patients who know their hemoglobin level.

Item response theory was used to shorten EORTC QLQ-C30 scales for use in palliative care

BACKGROUND AND OBJECTIVE

The goal was to develop a shortened version of the EORTC QLQ-C30 for use in palliative care. We wanted to keep as few items as possible in each scale while still being able to compare results with studies using the original scales. We examined the possibilities of shortening the physical functioning, cognitive functioning, fatigue, and nausea and vomiting scales.

STUDY DESIGN AND SETTING

The shortening was based on 2,366 (physical functioning) and 10,815 (three other scales) observations, respectively. We used item response theory to construct scoring algorithms for predicting scores on the original scales.

RESULTS

Evaluations showed that a three-item physical scale, a two-item fatigue scale, and a one-item nausea or vomiting scale predicted the scores on the original scales with excellent agreement and had measurement abilities similar to the original scales with no loss or only a little loss in power to detect group differences. The results of the cognitive functioning scale indicated problems when predicting scores from a shortened version.

CONCLUSION

Given the favorable results for the physical functioning, fatigue, and nausea or vomiting scales we expect that the shortened versions of these scales will be included in the abbreviated version of the EORTC QLQ-C30 for palliative care.

Patterns of drug prescription in a geriatric evaluation and management unit as compared with the general medical wards: a randomised study

OBJECTIVE

This study's objective was to determine whether patients treated in a geriatric evaluation and management unit (GEMU) had a more appropriate drug profile than patients treated in the general medical wards (MW).

METHODS

Frail elderly patients admitted as emergencies to the medical department were randomised to treatment in the GEMU (n=127) or MW (n=127). Drugs used at inclusion and discharge were registered retrospectively and analysed with regard to polypharmacy, number of drugs withdrawn or started, potential drug-drug interactions (DDIs), number of anticholinergic drugs prescribed, and the number of inappropriate drug prescriptions according to Beers' criteria. Utilisation of psychotropic and cardiovascular drugs was compared in detail according to prespecified hypotheses.

RESULTS

The number of patients with polypharmacy did not differ significantly between the GEMU and MW. The median number of scheduled drugs withdrawn per patient was higher in the GEMU than in the MW (p=0.005). Drugs with anticholinergic effects (p=0.003); cardiovascular drugs (p<0.001), particularly digitalis glycosides (p<0.001); and antipsychotic drugs (p=0.009) were withdrawn more often in the GEMU. The median number of scheduled drugs started was higher in the GEMU than in the MW (p=0.03). In particular, antidepressants (p<0.001) and estriol (p=0.001) were started more often in the GEMU than in the MW. Fewer GEMU than MW patients had potential DDIs at discharge (p=0.009).

CONCLUSION

Drug treatment in the GEMU as compared with the MW was more appropriate in terms of prescription of fewer drugs with anticholinergic effects and fewer potential DDIs. There were distinct differences in treatment patterns of cardiovascular and psychotropic drugs.

Scoring based on item response theory did not alter the measurement ability of EORTC QLQ-C30 scales

BACKGROUND AND OBJECTIVES

Most health-related quality-of-life questionnaires include multi-item scales. Scale scores are usually estimated as simple sums of the item scores. However, scoring procedures utilizing more information from the items might improve measurement abilities, and thereby reduce the needed sample sizes. We investigated whether item response theory (IRT)-based scoring improved the measurement abilities of the EORTC QLQ-C30 physical functioning, emotional functioning, and fatigue scales.

METHODS

Using a database of 13,010 subjects we estimated the relative validities of IRT scoring compared to sum scoring of the scales.

RESULTS

The mean relative validities were 1.04 (physical), 1.03 (emotional), and 0.97 (fatigue). None of these were significantly larger than 1. Thus, no gain in measurement abilities using IRT scoring was found for these scales. Possible explanations include that the items in the scales are not constructed for IRT scoring and that the scales are relatively short.

CONCLUSION

IRT scoring of the three longest EORTC QLQ-C30 scales did not improve measurement abilities compared to the traditional sum scoring of the scales.

Quality of life may be affected more by disease parameters and response to therapy than by haemoglobin changes

Earlier studies showing a negative impact of anaemia on quality of life (QOL) lack adequate adjustment for confounding factors such as disease stage and tumour response. We examined the impact of haemoglobin concentration on QOL scores of 745 multiple myeloma patients followed from diagnosis, adjusting for objective disease parameters. Data from two Nordic studies with the EORTC QLQ-C30 questionnaire were analysed using linear regression analysis. Haemoglobin was independently related only to fatigue at baseline (P = 0.001) and at 12 months (P = 0.010). In multivariate analysis, extent of skeletal disease was at least as strong a predictor for fatigue at diagnosis as haemoglobin and was also related to other important QOL scores such as physical functioning, role functioning, global QOL and pain (P < 0.001). At 12 months' follow-up, response to therapy was related to physical functioning (P < 0.001) and pain (P = 0.001). In conclusion, haemoglobin and extent of skeletal disease were both predictors for fatigue in patients with newly diagnosed multiple myeloma, but extent of skeletal disease was also associated with other important QOL scores. During follow-up, response to therapy emerged as an important predictor variable. When examining the effect of haemoglobin on QOL, it is essential to adjust for disease parameters and response to therapy in order not to overestimate the impact of haemoglobin on QOL. Our findings imply that uncontrolled studies on the effect of erythropoietin (EPO) in cancer patients may be making exaggerated claims for the effect of EPO on QOL.

Use of item response theory to develop a shortened version of the EORTC QLQ-C30 emotional functioning scale

BACKGROUND

As part of a larger study whose objective is to develop an abbreviated version of the EORTC QLQ-C30 suitable for research in palliative care, analyses were conducted to determine the feasibility of generating a shorter version of the 4-item emotional functioning (EF) scale that could be scored in the original metric.

METHODS

We used data from 24 European cancer studies conducted in 10 different languages (n = 8242). Item selection was based on analyses by item response theory (IRT). Based on the IRT results, a simple scoring algorithm was developed to predict the original 4-item EF sum scale score from a reduced number of items.

RESULTS

Both a 3-item and a 2-item version (item 21 'Did you feel tense?' and item 24 'Did you feel depressed?') predicted the total score with excellent agreement and very little bias. In group comparisons, the 2-item scale led to the same conclusions as those based on the original 4-item scale with little or no loss of measurement efficiency.

CONCLUSION

Although these results are promising, confirmatory studies are needed based on independent samples. If such additional studies yield comparable results, incorporation of the 2-item EF scale in an abbreviated version of the QLQ-C30 for use in palliative care research settings would be justified. The analyses reported here demonstrate the usefulness of the IRT-based methodology for shortening questionnaire scales.