Implementation of Computer-Based Quality-of-Life Monitoring in Brain Tumor Outpatients in Routine Clinical Practice

Article Outline

• Abstract
• Introduction
• Methods
  • Sample
  • Procedure
  • Assessment Instruments
  • QOL Monitoring—Computer-Based Data Assessment and Evaluation (CHES)
  • Training of the Medical Staff
  • Statistical Analysis
• Results
  • Patient Characteristics
  • Practical Experiences with the Implementation of QOL Monitoring into Clinical Routine
  • Reference Values for the EORTC QLQ-C30 and QLQ-BN20
  • Relevant Changes of QOL Scores
  • Individual CHES Patient Profile
    • Patient A.N.
• Discussion
• Acknowledgments
• References
• Copyright

Abstract 

Context

Computerized assessment of quality of life (QOL) in patients with brain tumors can be an essential part of quality assurance with regard to evidence-based medicine in neuro-oncology.

Objectives

The aim of this project was the implementation of a computer-based QOL monitoring tool in a neurooncology outpatient unit. A further aim was to derive reference values for QOL scores from the collected data to improve interpretability.

Methods

Since August 2005, patients with brain tumors treated at the neuro-oncology outpatient unit of the Innsbruck Medical University were consecutively included in the study. QOL assessment (European Organisation for Research and Treatment of Cancer [EORTC] Quality of Life Questionnaire [QLQ-C30] plus the EORTC QLQ-brain cancer module [BN20]) was computer-based, using a software tool called the Computer-based Health Evaluation System.

Results

A total of 110 patients with primary brain tumors (49% female; mean [standard deviation] age 47.9 [12.6] years; main diagnoses: 30.9% astrocytoma, 17.3% oligodendroglioma, 17.3% glioblastoma, 13.6% meningioma) was included in the study. On average, QOL was assessed 4.74 times per patient, 521 times in total. The user-friendly software was successfully implemented and tested. The routine QOL assessment was found to be feasible and was well accepted by both physicians and patients.

Conclusion

The software-generated graphic QOL profiles were found to be an important tool for screening patients for clinically relevant problems. Thus, computer-based QOL monitoring can contribute to an optimization of treatment (e.g., symptom management, psychosocial interventions) and facilitate data collection for research purposes.

Key Words: Computer, quality of life, assessment, monitoring, neuro-oncology

Back to Article Outline

Introduction 

Despite improving medical treatment options, brain tumors are known to have a bad prognosis, with five-year relative survival rates of about 34%.¹ Throughout the course of the disease, patients are confronted with tumor-related impairments in functioning and quality of life (QOL). Moreover, the multimodal treatment of the tumor is often accompanied by severe side effects that further reduce functioning and cause severe symptom burden itself.², ³ Knowledge of these impairments may enhance understanding of the disease and have an impact on its treatment. This points to the importance of integrating QOL assessment into clinical decision making and symptom management. Consequently, in recent years, QOL has been introduced as an outcome criterion that complements survival time. This has led to the integration of QOL assessment into treatment evaluation in oncology research, although much less often into clinical care.², ³, ⁴, ⁵, ⁶

Although the number of studies on QOL in patients with various cancers has been increasing, research on QOL in patients with brain tumors is still scarce. Studies on QOL in patients with brain tumors often comprise only a small number of subjects, are just descriptive and retrospective, or do not use adequate assessment instruments.

The relatively small number of studies of patients with brain tumors shows divergent results. Some studies demonstrate that the localization of the tumor is a determining factor of QOL, whereas others do not.⁷ Similarly, the impact of the histologic tumor type on QOL is controversial.³, ⁸ The meaning of other clinical parameters, such as tumor grading and treatment regimen (chemotherapy, radiotherapy) as well as sociodemographic variables (age, sex, and marital status) with regard to QOL, also appears to be unclear so far. Nevertheless, in studies evaluating chemotherapy regimens in patients with brain cancer, QOL assessments have been shown to provide important information on patient condition.⁹, ¹⁰, ¹¹ But QOL monitoring of patients with brain cancer in daily clinical routine is rarely conducted.

In this study, we wanted to present a prototypical implementation of computer-based QOL monitoring into the routine clinical practice of a neuro-oncology outpatient unit. Because interpretability of QOL data strongly depends on the availability of reference values from comparable samples, generation of such benchmarks was another focus of our study. As a software tool for the assessment and presentation of QOL data, we used the computer-based health evaluation system (CHES).¹², ¹³ Thus, our study addresses a range of different issues relevant for the implementation of QOL monitoring in clinical routine. In particular, the aims were as follows:

Back to Article Outline

Methods 

Sample 

Patients with primary brain tumors treated at the neuro-oncology outpatient unit of Innsbruck Medical University were considered for participation in our observational study. Inclusion criteria were diagnosis of a primary brain tumor, age between 18 and 80 years, fluency in German, no severe cognitive impairments, expected survival time of at least three months, attendance at the outpatient unit at least once per year, and informed consent. Reasons for dropout were recorded. In addition to patients' ratings, proxy ratings from significant others were collected (results will be published elsewhere). The study was approved by the local ethics committee.

Procedure 

After informed consent, the computerized versions of the EORTC QLQ-C30 and QLQ-BN20 were administered to the patients in the waiting room while they were waiting for their examinations at the neuro-oncology outpatient unit. Administration was done partly by a graduate psychology student doing an internship at the unit and partly by nurses.

A tablet personal computer (PC) presenting the questionnaires on its screen was handed over to the patients, along with instructions for the completion of the questionnaire and information on the purpose of the QOL monitoring. Time for completion of the questionnaire was recorded by the software.

Patients then filled in the questionnaire on the touch screen using a pen. After finishing, the patients gave the PC back to a member of the staff and went onto the medical examination. Subsequently, data were sent to a server database via a wireless local area network. With the help of the software CHES, QOL profiles were generated and immediately available to the physician as longitudinal charts (Fig. 1).

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1 

  QOL profile (bladder control and role functioning) for patient A.N.: aftercare (column 1–2), chemotherapy (column 3–10).

Assessment Instruments 

The EORTC QLQ-C30,¹⁴ an internationally validated and widely used cancer-targeted QOL instrument, assesses various facets of functioning, symptoms common in cancer patients, and global QOL. It has a modular structure consisting of a core questionnaire (EORTC QLQ-C30) and specific additional modules for cancer patients of different diagnostic groups.¹⁵

The core questionnaire consists of five functioning scales (physical functioning, social functioning, role functioning, emotional functioning, and cognitive functioning), a scale for global QOL and nine symptom scales (fatigue, nausea or vomiting, pain, dyspnea, sleeping disturbances, appetite loss, constipation, diarrhea, and financial impact). It has shown good psychometric performance and demands little time for completion (10–15 minutes). For the functioning scales and global QOL, high scores indicate good QOL, whereas high scores on the symptom scales reflect impaired QOL.

The additional brain cancer module (EORTC QLQ-BN20¹⁶) is a 20-item supplement for the QLQ-C30 that assesses brain cancer-specific QOL issues. The module comprises the subscales future uncertainty, visual disorders, bladder control, motor dysfunctions, headaches, communication deficits, seizures, hair loss, itchy skin, and weakness of legs. For these scales, high scores indicate low QOL. As a supplement, two items concerning taste and smell alteration were added from the EORTC item bank.

All scales were scored according to the EORTC guidelines, resulting in a score range from 0 to 100 points.

QOL Monitoring—Computer-Based Data Assessment and Evaluation (CHES) 

For the QOL monitoring, we used CHES,¹², ¹³ which was especially adapted for the present study. CHES is PC software for the computerized administration of questionnaires, scoring, and the graphical presentation of longitudinal data (questionnaires and medical data). The software presents scales as bar charts, together with reference lines indicating certain percentiles. Furthermore, a flag system was developed to mark those patients potentially in need of further medical and/or psycho-oncologic interventions. A red light means that a patient exceeds the 90th percentile, whereas an orange light is related to the 75th percentile. We preferred percentiles over means and standard deviations (SDs), as they provide more meaningful information for QOL data.¹⁷ Deterioration over time is represented by yellow triangles. Induced interventions (e.g., pain or fatigue intervention) also can be added to the graphical charts, allowing the tracking of QOL changes after specific interventions. The longitudinal graphical presentation, with its flag system, enables the physicians to detect QOL deficits at a single glance (Fig. 1).

Training of the Medical Staff 

The entire medical staff at the neuro-oncology outpatient unit of Innsbruck Medical University (two nurses and three physicians) agreed to participate in the study. Years of experience in the field of oncologic care ranged from two to 25.

In a one-hour training session, all staff members were trained in handling the software and interpreting the results from the EORTC QLQ-C30/BN20. Examples of individual QOL profiles and corresponding clinical data as well as adequate interventions were discussed. These training sessions were repeated every three months throughout the course of the study.

For interpretation of QOL profiles, we first used reference data from previous studies¹⁸, ¹⁹ and then switched to the reference values presented in this article.

Statistical Analysis 

Sample characteristics are presented as percentages, means, and SDs. For facilitating interpretation of QOL profiles, percentiles were calculated. To classify patients, we used the 10th, 25th, and 50th percentiles for the functioning scales and the 90th, 75th, and 50th percentiles for the symptom scales. This was done to identify patients with impairments of QOL on the basis of the distribution of an appropriate reference group.

Moreover, to classify QOL changes within individual subjects as “relevant,” we used a modification of an approach suggested by Crosby et al.²⁰ The proposed criterion is defined as the maximum of two thresholds, one governing clinical relevance and the other controlling statistical significance. A change exceeding one SD of the scale was defined as clinically relevant and a change larger than the 95% confidence interval of the test score was considered statistically significant. Details of the procedure are described elsewhere.²¹ SPSS 15.0(SPSS Inc., Chicago, Illinois) was used for all statistical analyses.

Back to Article Outline

Results 

Patient Characteristics 

Between May 2005 and November 2008, 157 patients with primary brain tumors at the neuro-oncology outpatient unit of Innsbruck Medical University were eligible for participation in the study. Forty-seven patients could not be included: 19 patients were in very bad physical condition (seven had severe cognitive impairments), 18 patients visited the outpatient unit less frequently than once per year, four patients did not provide informed consent, three patients were not fluent in German, and three patients had severe visual disorders. Thus, data from 110 patients were available for statistical analyses. Seventeen patients dropped out before the end of study because of their physical condition. Within the time of the project, 12 patients died.

Patients' QOL was assessed 521 times in total, an average of 4.74 times per patient. Mean patient age was 47.9 years (SD 12.6) and 50.1% were male. Sociodemographic and clinical data for the 110 participating patients at baseline are shown in Table 1.

Table 1. Descriptive Statistics for Sociodemographic and Clinical Data at Baseline (n=110)

Characteristic	Percentage
Age (yr)
Mean (SD)	47.9 (12.6)
Sex
Women	49.1
Men	50.9
Marital status
Single	12.3
Partnership, marriage	79.2
Divorced, separated	4.7
Widowed	3.8
Housing situation
Living alone	21.0
Living in partnership and/or with children	76.2
Living with family of origin	2.8
Education
Compulsory school	20.5
Apprenticeship, professional school	35.2
A-level	35.2
University degree	9.1
Employment status
Full time	28.3
Part time	7.6
Homemaker	5.4
Training	2.2
Retired	25.0
Status of employee's illness	18.5
Others	13.0
Duration of illness (mo)
Mean (SD)	65.0 (68.3)
Tumor type
Meningioma	13.6
Glioblastoma	17.3
Astrocytoma	30.9
Oligodendroglioma	17.3
Ependymoma	1.8
Embryonal tumors	4.5
Other	14.6
WHO grading
Grade I	18.8
Grade II	33.7
Grade III	27.7
Grade IV	19.8
Previous surgery
No surgery/biopsy	46.8
Partial resection	31.6
Total resection	21.6
Previous radiotherapy	65.7
Previous chemotherapy	57.8
Location of tumor
Right hemisphere	41.6
Left hemisphere	52.5
Both sides	5.9
Current treatment phase
Chemotherapy	19.7
Aftercare	80.3
Current chemotherapy regimen
Temozolomide	41.2
PCV	47.1
Depocyte	11.7

WHO=World Health Organization; PCV = Procarbazine, lomustine, and vincristine.

Practical Experiences with the Implementation of QOL Monitoring into Clinical Routine 

Computerized QOL monitoring in routine clinical practice showed high feasibility and required few resources in terms of time and manpower. At the first assessment, average total assessment time was about 10 minutes (including explanation and filling in of the questionnaire). Mean duration of questionnaire completion alone was 5.2 minutes (SD 6.6) for the QLQ-C30 and 3.1 minutes (SD 3.0) for the QLQ-BN20.

Total time for questionnaire completion decreased over the course of the study. For the fifth assessment, a patient needed on average 3.7 minutes (SD 3.0) for the QLQ-C30 and 2.1 minutes (SD 2.0) for the QLQ-BN20.

Patients' acceptance of the computerized questionnaires was high, resulting only in a few patients (n=4) refusing to participate in the study, that is, in the QOL monitoring. But a considerable number of patients (n=22) was not capable of filling in the questionnaires because of disease-related impairments (see the previous sections).

However, some of the patients included in the study were not capable of filling in the questionnaires on the touch screen at some point during the study (five patients showed mild cognitive impairments, four patients had forgotten their reading glasses, one patient was blind but asked for inclusion in the study, and one patient had impaired fine motor skills). For these 11 patients, the assessment was conducted as an interview at least at one assessment time point.

One patient raised severe concerns over data privacy but could be included in the study after comprehensive explanations. One patient was severely annoyed by the QLQ-BN20 item on future perspective and one patient indicated difficulties with changing item directions. As a benefit of the QOL monitoring, patients reported that the questions on QOL animated the subsequent discussion with the doctor. For the patients, it was often not evident that specific symptoms could be related to the brain tumor.

As expected, a crucial point was training the medical staff repeatedly in interpreting the QOL data, especially pointing out that global QOL alone is an insufficient measure of patients' physical and psychological status. Thus, the periodic interdisciplinary discussion about specific scales, their relation to other clinical data, and the initiation of adequate interventions were of vital importance.

Comments by the medical staff collected during the periodic training sessions indicated high user-friendliness. Only one nurse (with limited previous information technology experience) reported some difficulties regarding use of the software, in particular, regarding entry of longitudinal clinical data.

Overall benefit from computer-based QOL monitoring was considered high by two physicians and moderately high by one physician. They reported that the QOL monitoring contributed to better detection of symptoms. This was especially true for loss of bladder control, which often would have gone unnoticed without the QOL monitoring. Suggestions for further software improvement included the implementation of alert messages in the software (e.g., popup window) for patients with severe QOL impairments and the automatic generation of clinical reports on QOL (e.g., pdf documents).

Reference Values for the EORTC QLQ-C30 and QLQ-BN20 

To facilitate interpretation of QOL scores, reference values for the EORTC QLQ-C30 and QLQ-BN20 have been calculated from the data collected in the QOL monitoring. These reference scores (see Table 2, Table 3) comprise mean, SD, and the 10th, 25th, 50th, 75th, and 90th percentiles separately for brain cancer patients under chemotherapy (31 patients, 212 assessments) and in aftercare (93 patients, 309 assessments). Because data assessment was longitudinal, some patients were measured under chemotherapy and in aftercare. Data from each treatment phase were individually aggregated before calculating reference scores; hence, data from all assessment time points were included.

Table 2. EORTC QLQ-C30/BN20 Reference Values for Brain Cancer Patients Undergoing Chemotherapy

	Mean	SD	10%	25%	50%	75%	90%
EORTC QLQ-C30
Physical functioning	76.0	24.5	31.2	71.1	80.4	91.7	99.8
Social functioning	64.6	32.2	2.4	45.0	70.8	95.8	100.0
Role functioning	62.3	28.5	23.0	46.7	66.7	87.5	99.8
Emotional functioning	68.6	25.0	34.3	45.8	75.0	89.7	97.8
Cognitive functioning	67.8	30.1	7.9	45.0	79.2	92.9	98.4
Global QOL	57.2	20.0	25.7	45.8	57.8	70.0	84.2
Fatigue	38.5	26.4	2.3	11.1	40.7	55.6	77.0
Nausea/vomiting	13.4	16.7	0.0	0.0	6.7	20.4	33.3
Pain	12.8	16.9	0.0	0.0	6.7	18.8	37.8
Dyspnea	14.7	20.4	0.0	0.0	3.3	22.2	47.1
Sleeping disturbances	22.3	24.8	0.0	0.0	13.3	44.4	65.3
Appetite loss	21.5	26.1	0.0	0.0	11.1	42.9	66.7
Constipation	16.9	21.8	0.0	0.0	8.3	33.3	41.3
Diarrhea	8.7	15.8	0.0	0.0	0.0	13.3	32.4
Financial impact	19.4	24.0	0.0	0.0	6.7	40.0	52.7
Taste alterations	22.6	27.9	0.0	0.0	8.9	33.3	79.6
EORTC QLQ-BN20
Future uncertainty	31.3	28.7	1.2	8.3	19.8	55.0	81.7
Visual disorders	16.0	20.0	0.0	0.0	6.3	24.4	53.6
Motor dysfunction	19.0	20.9	0.0	1.9	13.3	25.4	59.1
Communication deficits	20.6	25.0	0.0	0.0	8.6	33.3	63.9
Headaches	19.8	27.1	0.0	0.0	13.3	26.7	63.3
Seizures	10.0	23.3	0.0	0.0	0.0	8.9	53.5
Drowsiness	33.4	29.8	0.0	13.3	27.8	51.1	94.0
Hair loss	13.0	26.8	0.0	0.0	2.2	11.1	60.0
Itchy skin	11.0	21.9	0.0	0.0	0.0	13.3	41.0
Weakness of legs	28.6	29.9	0.0	0.0	18.5	60.0	74.3
Bladder control	6.4	14.1	0.0	0.0	0.0	3.7	30.5

Note: 31 patients, 212 assessments. Bold values indicate clinically relevant percentiles.

Table 3. EORTC QLQ-C30/BN20 Reference Values for Brain Cancer Patients in Aftercare

	Mean	SD	10%	25%	50%	75%	90%
EORTC QLQ-C30
Physical functioning	83.5	19.5	56.2	70.4	90.8	100.0	100.0
Social functioning	75.9	24.1	41.3	61.1	81.9	100.0	100.0
Role functioning	72.8	27.9	33.3	52.8	83.3	100.0	100.0
Emotional functioning	68.5	20.7	40.0	52.8	66.7	89.4	97.2
Cognitive functioning	71.6	27.3	33.3	50.0	83.3	95.5	100.0
Global QOL	65.8	19.9	41.7	52.1	66.7	79.6	95.0
Fatigue	33.2	23.8	0.0	11.1	33.3	50.0	66.7
Nausea/vomiting	7.5	12.2	0.0	0.0	0.0	15.5	23.3
Pain	23.0	27.7	0.0	0.0	11.1	34.7	66.7
Dyspnea	15.7	21.4	0.0	0.0	0.0	26.9	46.7
Sleeping disturbances	25.3	30.2	0.0	0.0	11.1	43.1	72.4
Appetite loss	11.9	18.9	0.0	0.0	0.0	22.2	33.3
Constipation	13.1	20.7	0.0	0.0	0.0	19.4	44.4
Diarrhea	10.2	18.8	0.0	0.0	0.0	16.7	33.3
Financial impact	18.4	27.5	0.0	0.0	0.0	33.3	66.7
Taste alterations	9.4	17.9	0.0	0.0	0.0	16.7	36.0
EORTC QLQ-BN20
Future uncertainty	25.5	21.8	0.0	6.3	20.8	41.7	56.7
Visual disorders	11.2	17.1	0.0	0.0	3.7	17.1	33.3
Motor dysfunction	15.9	18.4	0.0	0.0	11.1	25.9	44.4
Communication deficits	20.8	23.0	0.0	0.0	14.8	33.3	57.2
Headaches	30.3	30.5	0.0	0.0	25.0	50.0	81.1
Seizures	10.0	21.5	0.0	0.0	0.0	1.1	47.8
Drowsiness	28.7	25.1	0.0	0.0	33.3	50.0	66.7
Hair loss	7.3	17.0	0.0	0.0	0.0	0.0	33.3
Itchy skin	10.9	23.3	0.0	0.0	0.0	11.1	43.3
Weakness of legs	16.2	26.1	0.0	0.0	0.0	31.9	66.7
Bladder control	9.6	19.2	0.0	0.0	0.0	11.1	33.3

Note: 93 patients, 309 assessments. Bold values indicate clinically relevant percentiles.

We consider the 10th percentile (for functioning scales) and the 90th percentile (symptom scales), respectively, as being of particular clinical relevance because patients exceeding these scores are likely to be in need of additional care.

Relevant Changes of QOL Scores 

Thresholds for relevant changes for each multi-item scale of the QLQ-C30 and QLQ-BN20 are shown in Table 4. Because of the discreteness of the subscales, only a limited number of distinct change scores can occur. For this reason, thresholds were rounded to the closest discrete score. This rounding led to the leveling of small discrepancies between the thresholds for clinical relevance and statistical significance.

Table 4. Clinical Relevance and Statistical Significance of Changes in QLQ-C30 Scores and QLQ-BN20 Scores

	Thresholds for Clinical Relevance and Statistical Significance and Their Combinations
	Threshold for Clinical Relevance (C) “1 SD” Exact/In Effecta	Threshold for Stat. Significance (S) “MDC” Exact/In Effecta	Threshold for Combined Criteria: Max (1 SD, MDC) Exact/In Effecta	Aspect Determining the Threshold of the Combined Criteriab	% “Relevant” Changes (Among All Changes)
EORTC QLQ-C30
Physical functioning	20.8/20.0c	20.3/20.0	20.8/20.0	C, S likewise	14.2
Role functioning	29.2/33.3	33.1/33.3	33.1/33.3	C, S likewise	29.5
Emotional functioning	27.2/25.0	28.7/25.0	28.7/25.0	C, S likewise	29.2
Social functioning	30.2/33.3	30.6/33.3	30.6/33.3	C, S likewise	25.7
Cognitive functioning	29.6/33.3	35.8/33.3	35.8/33.3	C, S likewise	17.9
Global QOL	23.3/25.0	19.4/16.7	23.3/25.0	C	22.3
Fatigue	27.4/22.2	29.7/33.3	29.7/33.3	S	21.8
Pain	27.9/33.3	32.1/33.3	32.1/33.3	C, S likewise	21.8
Nausea or vomiting	18.8/16.7	31.4/33.3	31.4/33.3	S	18.4
Taste alterations	27.2/33.3	23.4/16.7	27.2/33.3	C	22.5
EORTC QLQ-BN20
Future uncertainty	27.0/25.0	27.6/25.0	27.6/25.0	C, S likewise	25.9
Visual disorders	17.6/22.2	24.3/22.2	24.3/22.2	C, S likewise	19.1
Motor dysfunction	19.8/22.2	33.4/33.3	33.4/33.3	S	12.2
Communication deficits	25.8/22.2	24.9/22.2	25.8/22.2	C, S likewise	29.4

After adjusting to discrete scores, the thresholds for clinical relevance and statistical significance were identical for nine of 14 scales. For global QOL and taste alterations, clinical relevance thresholds were higher, whereas for pain, nausea or vomiting, and motor dysfunction, thresholds for statistical significance were higher. In case of dissimilarity, the higher threshold was used.

Thresholds ranged from 20.0 (physical functioning) to 33.3 (eight scales). The corresponding percentage of relevant changes among all changes was lowest for physical functioning (14.2%) and highest for emotional functioning (29.5%).

Individual CHES Patient Profile 

To illustrate data from QOL monitoring in more detail, the following section presents an example for an individual course of QOL.

The male patient A.N. was born in 1944 and was diagnosed with a glioblastoma multiforme (right temporal) in December 2003 and underwent surgery with partial resection and subsequent radiotherapy and chemotherapy with temozolomide. In June 2004, further surgery was necessary and he received chemotherapy with procarbazine, lomustine, and vincristine from July 2004 until May 2005. In May 2005, he was included in the QOL monitoring. Figure 1 shows the course of bladder control problems (EORTC QLQ-BN20) and role functioning (EORTC QLQ-C30) from October 2006 till May 2007. In November 2006, chemotherapy with lomustine was started.

As can be seen in Fig. 1, the patient's role functioning deteriorated over the course of chemotherapy, being mostly below the 10th percentile of the reference group. Bladder control also was strongly impaired at the beginning of chemotherapy and again at the end of chemotherapy. The deterioration of bladder control between the last two assessment time points proved to represent relevant change.

Back to Article Outline

Discussion 

Monitoring of QOL in routine clinical practice should be considered as an integral part of oncologic care, which may contribute essentially to an individualization of treatment. Because patient-physician contacts in busy outpatient units are rather short, many impairments may not be detected, especially if frequency of appointments is low. Furthermore, evaluating the long-term course in symptom burden and functioning, as well as tracking consequences of specific interventions, may prove difficult without standardized, quantitative assessment instruments.

Our study investigated the feasibility of standardized detailed long-term evaluation of QOL in patients with brain tumors treated in a neuro-oncology outpatient unit. The software support provided by CHES enabled the collection and processing of QOL data.

Because CHES meets requirements regarding usability, most of the patients were able to complete the questionnaire on their own and only a very few patients needed help completing the questionnaire. In general, the computerized QOL monitoring proved feasible with respect to patients, the participating proxy raters, and the physicians. For patients with severe visual or motor impairments, adequate technological solutions are planned for future monitoring procedures. In spite of the user-friendliness of the software and the low degree of help needed, the computerized data collection needs to be supervised in the daily routine to guarantee gapless administration and to provide a contact person for possible questions concerning the questionnaires or the monitoring itself. In longitudinal assessments, this might be relevant mainly at the first assessment time point. At subsequent assessment time points, patients need less help, as indicated by the considerable decrease of assessment time in our study. In comparison to information in the EORTC QLQ-C30 manual,²² the duration of filling in the questionnaire was much shorter in our study. This might be attributed to the use of a computerized questionnaire version being more legible and better arranged than a paper-pencil version.

On the part of the physicians, the implementation of routine QOL evaluation and its use depends on their attitude toward QOL projects and also on their “computer friendliness.” At the neuro-oncology outpatient unit, the computer-based QOL monitoring could be successfully implemented. An important reason for this was the high commitment of all team members. For the medical staff, training in the interpretation of the QOL profiles turned out to be an important aspect. This was taken into account by periodic training sessions comprising discussions focusing on QOL data of individual subjects. This improves correct interpretation of the data, including longitudinal charts and the interaction of different symptoms and functioning scales.

A crucial issue with regard to interpretation of QOL data is the determination of reference values, such as cut-off values or percentiles. Clear recommendations on how to use reference values for the EORTC questionnaires are missing in the literature. The manual for the EORTC QLQ-C30²³ offers score distributions in various clinical groups as a basis for interpretation of these scores but does not contain cut-off values. An approach referring to general population distributions was suggested by Fayers¹⁷ who proposed the 10th percentile as a useful reference point. However, taking this lack of clear guidelines into account, we chose a provisional distribution-based approach for our study, using percentiles from brain cancer patients' on- and off-treatment. Further research is needed on this issue, especially on adaptation of reference values to certain clinical and sociodemographic variables (e.g., diagnosis, treatment phase, age, and sex). A criterion-based approach to determination of cut-off values poses the problem of finding appropriate criteria, which might be feasible for scales such as physical functioning or diarrhea, but much more difficult for scales such as pain or global QOL. Furthermore, the interpretation of longitudinal data may be affected by matters of response shift biasing the course of QOL scores over time.

Nevertheless, in our clinical setting, routine QOL evaluation proved to be a useful tool for focusing on patients' needs and it can be easily included into routine clinical practice.

Similar results also were found in other studies on the implementation of routine computer-based QOL monitoring. In a randomized controlled trial, Velikova et al.²⁴ found QOL monitoring to be a feasible approach for improving medical practice. In their study, QOL monitoring had a positive effect on emotional well-being and improvement of overall QOL. In the patient group under monitoring, they found a more frequent discussion of chronic nonspecific symptoms without prolonging encounters. Improvement of overall QOL was associated with explicit use of QOL data, discussion of pain, and role functioning.

For the collection of QOL data in clinical routine, it is considered as crucial that the evaluation procedure does not interfere with the workflow of a busy outpatient unit.²⁵ Many of the practical problems associated with measuring QOL in clinical practice are assumed to be overcome by the use of new technologies.

The feasibility of implementation of computer-based QOL assessment also was proven in several other studies.²⁶, ²⁷, ²⁸, ²⁹ Especially, the immediate presentation of results to clinicians informing the treating physician about important issues with no time loss was described as crucial.²⁴, ³⁰, ³¹, ³² Thus, QOL evaluation is not an extra burden for the patients but is considered to be an integral part of the individual treatment.

Future improvement of QOL monitoring is expected from notebook-based ward rounds and integration of QOL findings into clinical information systems. Beyond its relevance for clinical practice, QOL data acquired through routine monitoring provides a comprehensive data basis for research purposes.

We expect that the use of the QOL profiles in daily clinical routine has the potential of changing clinical practice and can be a basis for the initiation of medical and psycho-oncologic interventions. The experiences of this study indicate that the computer-based routine QOL assessment is useful for early detection of physical symptoms and psychosocial problems. However, at present, the precise implications of QOL monitoring on everyday clinical practice in the neuro-oncology outpatient unit are not known in detail, because this study focused on implementation issues rather than on evaluation. A further study assessing the impact of QOL data on physician-patient communication and treatment decision is planned.

Back to Article Outline

Acknowledgments 

The authors want to thank Jakob Pinggera, Stefan Zugal, and Barbara Weber for help with software programming. In addition, they want to thank Elisabeth Huber and Theresia Kindl for help with data collection.

Back to Article Outline

References 

1. National Cancer Institute . Surveillance epidemiology and end results (SEER) cancer statistics review, 1975-2005. Bethesda, MD: National Cancer Institute; 2008;Available from http://seer.cancer.gov/csr/1975_2005Accessed October 17, 2008
   • View In Article
2. Efficace F, Bottomley A. Health related quality of life assessment methodology and reported outcomes in randomised controlled trials of primary brain cancer patients. Eur J Cancer. 2002;38:1824–1831
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (109 KB)
   • CrossRef
3. Hahn C, Dunn R. Prospective study of neuropsychologic testing and quality of life assessment of adults with primary malignant brain tumors. Int J Radiat Oncol Biol Phys. 2003;55(4):992–999
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (85 KB)
   • CrossRef
4. Lahrmann H, Oberndorfer S, Grisold W. Therapeutic concepts for primary malign tumors. Journal für Neurologie, Neurochirurgie und Psychiatrie. 2001;2(2):16–20
   • View In Article
5. Higginson IJ, Carr AJ. Measuring quality of life: using quality of life measures in the clinical setting. BMJ. 2001;322(7297):1297–1300
   • View In Article
   • MEDLINE
   • CrossRef
6. Middeke M, Bauhofer A, Kopp I, Koller M. Computerized visualization of quality of life data of individual cancer patients—the QoL-Profiler. Inflamm Res. 2004;53(Suppl 2):175–178
   • View In Article
7. Palese A, Lamanna F, Di-Monte C, et al. Quality of life in patients with right- or left-sided brain tumors: literature review. J Clin Nurs. 2008;17(11):1403–1410
   • View In Article
   • CrossRef
8. Salo J, Niemelä A, Joukamaa M, Koivukangas J. Effect of brain tumor laterality on patient's perceived quality of life. J Neurol Neurosurg Psychiatry. 2002;72:373–377
   • View In Article
   • MEDLINE
   • CrossRef
9. MacDonald D, Kiebert G, Padros M, Yung A, Olson J. Benefit of temozolomide compared to procarbazine treatment of glioblastoma multiforme at first relapse: effect on neurological functioning, performance status, and health related quality of life. Cancer Invest. 2005;23(2):138–144
   • View In Article
   • MEDLINE
   • CrossRef
10. Prados M, Schold S, Fine H, et al. A randomized, double-blind, placebo-controlled, phase 2 study of RMP-7 in combination with carboplatin administered intravenously for the treatment of recurrent malignant glioma. Neuro Oncol. 2003;5(2):96–103
    • View In Article
    • MEDLINE
    • CrossRef
11. Hart M, Grant R, Garside R, et al. Temozolomide for high grade glioma. Cochrane Database Syst Rev. 2008;(4):CD007415
    • View In Article
12. Holzner B, Zabernigg A, Kemmler G, et al. Computerized assessment of quality of life in patients undergoing chemotherapy. Qual Life Res. 2004;13(9):1523
    • View In Article
13. Holzner B, Giesinger J, Rumpold G, Weber B, Kemmler G. Computerized assessment of quality of life in daily clinical routine. EORTC Quality of Life Group Newsletter 2008;(7).
    • View In Article
14. Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst. 1993;85(5):365–376
    • View In Article
    • MEDLINE
    • CrossRef
15. Sprangers MA, Cull A, Groenvold M, et al. The European Organization for Research and Treatment of Cancer approach to developing questionnaire modules: an update and overview. EORTC Quality of Life Study Group. Qual Life Res. 1998;7(4):291–300
    • View In Article
    • MEDLINE
    • CrossRef
16. Osoba D, Aaronson NK, Muller M, et al. The development and psychometric validation of a brain cancer quality-of-life questionnaire for use in combination with general cancer-specific questionnaires. Qual Life Res. 1996;5(1):139–150
    • View In Article
    • MEDLINE
    • CrossRef
17. Fayers PM. Interpreting quality of life data: population-based reference data for the EORTC QLQ-C30. Eur J Cancer. 2001;37(11):1331–1334
    • View In Article
    • Full Text
    • Full-Text PDF (143 KB)
    • CrossRef
18. Taphoorn MJ, Stupp R, Coens C, et al. Health-related quality of life in patients with glioblastoma: a randomised controlled trial. Lancet Oncol. 2005;6:937–944
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (146 KB)
    • CrossRef
19. Osoba D, Brada M, Prados MD, Yung WK. Effect of disease burden on health-related quality of life in patients with malignant gliomas. Neuro Oncol. 2000;2(4):221–228
    • View In Article
    • MEDLINE
    • CrossRef
20. Crosby R, Kolotkin R, Williams G. An integrated method to determine meaningful changes in health-related quality of life. J Clin Epidemiol. 2004;57(11):1153–1160
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (207 KB)
    • CrossRef
21. Kemmler G, Zabernigg A, Gattringer K, et al. A new approach to combining clinical relevance and statistical significance for evaluation of quality of life changes in the individual patient. J Clin Epidemiol. July 15 2009;[Epub ahead of print]
    • View In Article
22. Young T, De Haes H, Curran D, et al. Guidelines for assessing quality of life in EORTC clinical trials. Brussels, Belgium: EORTC; 1999;
    • View In Article
23. Scott N, Fayers P, Aaronson N, et al. Group obotEQoL EORTC QLQ-C30 reference values. Brussels, Belgium: EORTC Quality of Life Group; 2008;
    • View In Article
24. Velikova G, Booth L, Smith AB, et al. Measuring quality of life in routine oncology practice improves communication and patient well-being: a randomized controlled trial. J Clin Oncol. 2004;22(4):714–724
    • View In Article
    • CrossRef
25. Buxton J, White M, Osoba D. Patients' experiences using a computerized program with a touch-sensitive video monitor for the assessment of health-related quality of life. Qual Life Res. 1998;7(6):513–519
    • View In Article
    • MEDLINE
    • CrossRef
26. Taenzer PA, Speca M, Atkinson MJ, et al. Computerized quality-of-life screening in an oncology clinic. Cancer Pract. 1997;5(3):168–175
    • View In Article
    • MEDLINE
27. Carlson LE, Speca M, Hagen N, Taenzer P. Computerized quality-of-life screening in a cancer pain clinic. J Palliat Care. 2001;17(1):46–52
    • View In Article
    • MEDLINE
28. Wright EP, Selby PJ, Crawford M, et al. Feasibility and compliance of automated measurement of quality of life in oncology practice. J Clin Oncol. 2003;21(2):374–382
    • View In Article
    • CrossRef
29. Cull A, Gould A, House A, et al. Validating automated screening for psychological distress by means of computer touchscreens for use in routine oncology practice. Br J Cancer. 2001;85(12):1842–1849
    • View In Article
    • MEDLINE
    • CrossRef
30. Chang CH, Cella D, Masters GA, et al. Real-time clinical application of quality-of-life assessment in advanced lung cancer. Clin Lung Cancer. 2002;4(2):104–109
    • View In Article
    • Abstract
    • Full-Text PDF (833 KB)
    • CrossRef
31. Fortner B, Okon T, Schwartzberg L, Tauer K, Houts AC. The Cancer Care Monitor: psychometric content evaluation and pilot testing of a computer administered system for symptom screening and quality of life in adult cancer patients. J Pain Symptom Manage. 2003;26(6):1077–1092
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (163 KB)
    • CrossRef
32. Fortner B, Baldwin S, Schwartzberg L, Houts AC. Validation of the Cancer Care Monitor items for physical symptoms and treatment side effects using expert oncology nurse evaluation. J Pain Symptom Manage. 2006;31(3):207–214
    • View In Article