Fax: (011) 31 20 512 2322
Use of health-related quality-of-life assessments in daily clinical oncology nursing practice
A community hospital-based intervention study
Article first published online: 9 JUN 2008
Copyright © 2008 American Cancer Society
Volume 113, Issue 3, pages 628–637, 1 August 2008
How to Cite
Hilarius, D. L., Kloeg, P. H., Gundy, C. M. and Aaronson, N. K. (2008), Use of health-related quality-of-life assessments in daily clinical oncology nursing practice. Cancer, 113: 628–637. doi: 10.1002/cncr.23623
- Issue published online: 18 JUL 2008
- Article first published online: 9 JUN 2008
- Manuscript Accepted: 31 MAR 2008
- Manuscript Revised: 6 MAR 2008
- Manuscript Received: 13 NOV 2007
- health-related quality of life;
- clinical practice;
- oncology nursing;
- community hospital;
- outpatient chemotherapy;
The current study evaluated the efficacy of incorporating standardized health-related quality of life (HRQL) assessments as a routine part of the outpatient chemotherapy treatment of cancer patients in a community hospital in terms of: 1) facilitating nurse-patient communication, 2) increasing nurses' awareness of patients' HRQL, 3) patient management, 4) patients' satisfaction, and 5) patients' HRQL.
The study employed a sequential cohort design with repeated measures. Ten nurses and 219 patients participated in this community hospital-based study. The intervention involved patients completing standardized HRQL questionnaires via a touch-screen computer, the results of which were provided to nurses and patients in a graphic summary. Questionnaire and medical record data were used to assess outcomes.
HRQL-related topics were discussed significantly more frequently in the intervention group than in the control group (mean = 4.8 topics vs 3.8 topics, respectively; P = .02). Nurses' awareness of patients' levels of daily activity, pain, and overall quality of life was significantly better in the intervention than the control group. The mean number of HRQL-related notations in the medical records was significantly higher in the intervention group (24 vs 20; P< .05). Only modest effects were observed in patient management (counseling behavior), and no significant effects were found in patient satisfaction or changes in HRQL over time.
Incorporating standardized HRQL assessments in daily clinical oncology nursing practice primarily facilitates the discussion of HRQL issues and increases nurses' awareness. Additional efforts are needed to enhance the effect of routine HRQL assessments on patient management and HRQL. Cancer 2008. © 2008 American Cancer Society.
There has been growing interest in the use of health-related quality of life (HRQL) assessments in daily clinical practice. Typically, this involves asking patients to complete a questionnaire at the time of an outpatient visit and generating a summary of the results that is then made available to the healthcare provider immediately before the consultation. It has been hypothesized that the availability of such standardized HRQL data can have a cascade of effects; 1) including enhanced provider-patient communication; 2) improved provider awareness/detection of patients' problems; 3) greater patient and provider satisfaction; 4) improved patient management; and, ultimately; 5) improved HRQL.1
This putative cascade effect has been investigated in several studies in the clinical oncology setting.1–4 In 3 of these studies, the provision of structured HRQL information was found to facilitate physician-patient communication.1, 3, 4 Two studies documented an increase in physicians' awareness of their patients' physical and psychosocial well-being.1, 4 One study demonstrated changes in patient management, and another observed a positive effect on limited aspects of the patients' HRQL.1, 4 All of these studies employed an HRQL questionnaire developed for use among cancer patients in general (the European Organization for Research and Treatment of Cancer [EORTC] QLQ-C30). There is increasing evidence that more detailed, condition-specific questionnaires may have value added beyond the more generic class of measures.5–7
Although the physician remains the central figure in the treatment of cancer patients, the role of the oncology nurse has increased substantially in recent years, particularly in the outpatient setting. In addition to their responsibility for obtaining patient histories, performing physical examinations, and delivering chemotherapeutic treatment, oncology nurses typically have more in-depth, day-to-day contact with the patient than does the medical oncologist. This often involves monitoring of patients' physical symptoms and their levels of physical and psychosocial functioning. Oncology nurses also play an important role as liaison between the patient and the oncologist, and as such can help to ensure that HRQL-related issues are addressed.8–11 In a small study directed at oncology nurses rather than physicians, the provision of structured HRQL information was associated with less symptom distress over time.12 However, a more recent study reported that HRQL assessment followed by a structured interview and discussion with a nurse did not improve patients' HRQL or satisfaction over time.13 Neither of these studies assessed the impact of standardized HRQL assessment on communication, awareness levels, or on patient management.
Finally, all studies but 112 were conducted in specialized cancer treatment centers or university hospitals. The question remains as to whether routine use of HRQL assessments can be incorporated successfully in community hospitals, in which large numbers of cancer patients receive treatment.
In the current study, we investigated the impact of providing oncology nurses with timely, structured feedback on the general and condition-specific HRQL of patients receiving outpatient chemotherapy. Our primary hypothesis was that such a procedure would enhance HRQL-related nurse-patient communication. Additional outcomes included nurses' awareness of their patients' HRQL, clinical management activities, patients' satisfaction, and patients' self-reported HRQL over time.
MATERIALS AND METHODS
Study Site and Subjects
The study was conducted at the Medical Center Alkmaar, a general community hospital located north of Amsterdam with approximately 1700 newly diagnosed cancer patients per year. The study sample included all of the oncology nurses working in the outpatient chemotherapy clinic and a consecutive series of patients who were to begin either adjuvant or palliative chemotherapy treatment. Patients were excluded if they lacked basic proficiency in Dutch, were aged <18 years, exhibited overt psychopathology or serious cognitive problems, or were participating in a concurrent HRQL study. Eligible patients were invited to participate in the study after they had received 1 cycle of chemotherapy. They were provided with both oral and written information before obtaining their informed consent. The Institutional Review Board of the hospital approved the study.
The study employed a sequential cohort design with repeated measures. An initial cohort of 100 consecutive patients formed the usual care control arm. After a ‘wash out’ period of approximately 2 months, a second cohort of consecutive patients formed the experimental arm of the study. Both patient cohorts were followed for 4 consecutive outpatient visits.
This study design was deemed the most appropriate and feasible, given the practical and ethical constraints involved. A randomized clinical trial (RCT) would have required that, during the same time period, the participating nurses be exposed to the experimental intervention for some patients, but not for others. In such a situation the risk of a carryover or contamination effect was considered too great. Similarly, an RCT would require that, during the same time period, some patients be exposed to the intervention whereas others would not. Given the physical layout of the clinic (ie, open waiting room areas, multipatient treatment rooms), this would likely have lead to confusion and between-patient contamination.
Baseline assessments took place at the second outpatient visit (first study visit) for both groups. The intervention took place during the third through the fifth outpatient visit for patients in the experimental condition. Outcome assessments took place at the fifth outpatient visit (fourth study visit) for all patients.
After the first and the fourth study visits all patients completed a set of questionnaires assessing HRQL-related communication with the oncology nurse, their HRQL, and their satisfaction with the quality of nursing care. The questionnaire assessing HRQL-related communication was completed at the outpatient clinic immediately after the chemotherapy treatment. All other questionnaires were completed at home and returned by mail to the study coordinating office. At the fourth study visit, patients in the experimental arm and the nurses were asked to rate the usefulness of the intervention. After completion of the first and fourth study visits, the nurses were asked to complete a questionnaire assessing the HRQL of the patient.
Notations relating to the patients' HRQL (functional problems and symptoms) and subsequent actions taken during the study period were extracted from the medical records.
Description of the Intervention
All patients had standard consultations with their physicians and nurses during the course of treatment. Nurses were responsible for the day-to-day care of the patients and consulted the treating physicians when necessary. Patients in the intervention arm were asked to complete a standardized HRQL questionnaire, the EORTC QLQ-C30 (version 3.0) and, if applicable, a condition-specific EORTC module assessing HRQL-related problems in breast cancer (the QLQ-BR23), colorectal cancer (the QLQ-CR38), or lung cancer (the QLQ-LC13).5–7, 14
Patients completed the questionnaires at computer workstations installed in the outpatient clinic, using Anyquest touch-screen software.15 The questionnaire results were summarized in graphic form (Fig. 1). Copies of the summary were given to the patient and nurse immediately before the consultation. A copy was also placed in the medical records. At study visits 3 and 4, the HRQL summary included both the patients' current scores and those elicited at the previous visit(s). Before the start of the intervention, the nurses received a group educational session and written information on how to interpret the HRQL summary scores. It was explained that the summary profiles could assist the nurses during the consultation with the patients, but that it was not obligatory to address all topics included in the summary. It was explained that a difference in scores between 2 visits of ≥10 points was indicative of a potentially meaningful improvement or deterioration in symptoms or functioning. Patients in the intervention group received similar written materials and, if desired, further explanation from a research assistant. No specific guidelines were provided to either the nurses or the patients regarding how the HRQL summary data could or should be used during the consultations.
Patients' sociodemographic and clinical characteristics
Age, education, and marital status were obtained from the patient questionnaires. Data on primary diagnosis and nature of chemotherapy (curative/adjuvant vs palliative) were extracted from the medical records.
Immediately after the first and the last study visit, patients were asked to report the specific HRQL topics covered by the QLQ-C30 and, when applicable, by the condition-specific module that had been discussed (dichotomous assessment; yes/no) with the nurse.
This self-report questionnaire has been validated previously and yields both a composite index (ie, the total number of HRQL-related topics discussed) and individual item scores (ie, per HRQL topic). In the validation study,16 the mean percentage agreement between patients' self-report versus observational data on topics discussed was 86%, with a mean prevalence-adjusted Cohen kappa of 0.71.
Nurses' awareness of patients' HRQL
At the first and last study visits, both the patients and the nurses completed the Dartmouth Primary Care Cooperative Information Functional Health Assessment (COOP) and the World Organization Project of National Colleges and Academics (WONCA) charts.17 These charts assess physical fitness, feelings, daily and social activities, overall health, pain, fatigue, and quality of life. A 5-point response scale is used, with 1 representing the best and 5 representing the worst level of functioning.
All notations relating to HRQL-related topics covered by the QLQ-C30 and modules (ie, symptoms or functional problems) were abstracted from the medical and nursing records, as were any patient management activities related to those topics (ie, prescription of medication, consultation with treating oncologist, ordering of tests, referrals to other healthcare providers, and counseling/advice-giving). Data were abstracted using a checklist (with ‘yes’ vs ‘no’ rating options). Two research assistants independently reviewed the charts and completed the checklist. Discrepant ratings occurred in <10% of the cases. All such discrepancies were resolved via discussion between the raters, in some cases with the researcher as arbiter.
Two composite scores were calculated by summing the HRQL-related notations and patient management actions, respectively.
After the first and the last study visits, patients completed a modified version of the patient satisfaction questionnaire developed for use in the outpatient clinics of the Netherlands Cancer Institute based, in part, on the item/scale content of the Patient Satisfaction Questionnaire (PSQ, Form II).18 The core of the questionnaire is composed of 20 items organized into 4 subscales: 1) perceived technical quality of care, 2) interpersonal manner, 3) communication, and 4) continuity of care. The modification of the questionnaire involved rewording the core set of questions such that the focus of evaluation was on nurses rather than physicians.
Patients' self-reported HRQL
After the first and last study visits, all patients completed the generic Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36).19, 20 The SF-36 is organized into 8 multi-item scales. In addition, the Functional Assessment of Cancer Therapy Breast Cancer (FACT-BCS),21 Colorectal Cancer (FACT-C),22 and Lung Cancer (FACT-L)23 questionnaires were administered to patients with breast, colon, and lung cancer, respectively. These questionnaires were used rather than the EORTC HRQL questionnaires to avoid contamination of the outcome assessment of changes in HRQL over time by the intervention tool itself (ie, patients in the intervention group had completed the EORTC questionnaire repeatedly as part of the intervention, whereas patients in the control group had not).
Patient and nurse evaluation of the intervention
After the last study visit, both patients in the intervention group and the nurses completed a questionnaire regarding their experience with the intervention.
Possible differences in the background characteristics of the patients in the intervention and control groups were tested using the chi-square statistics for categoric data, the Mann-Whitney statistic for ordinal data, and the Student t test for interval level data.
All statistical analyses relating to the effectiveness of the intervention were performed on an intention-to-treat basis and were based on data obtained from the fourth study visit. Patients' sex and baseline (ie, first visit) values were used as covariates. Linear regression analysis was used to test for between-group differences in the mean scores on the composite communication scale, patient satisfaction measures, and patients' self-reported HRQL. Logistic regressions analysis was used to compare the percentage of consultations in the intervention versus control group in which specific HRQL-related topics were discussed. Nurses' awareness of patients' HRQL problems was assessed by calculating intraclass correlation coefficients (ICCs) between patients' and nurses' ratings on the COOP/WONCA charts.24 Poisson regression was used to test for between-group differences in patient management outcomes.
The sample size was established to provide 80% power to detect a moderate between-group difference (effect size = 0.40) in the frequency with which patients' problems were discussed during the consultations, with α set at 0.05.
Nurse and Patient Recruitment and Sample Description
All 10 oncology nurses asked to participate in the study agreed to do so. All nurses were female and their mean age was 36 years (range, 26-48 years), with an average of 9 years of experience in oncology (range, 1-19 years). Throughout the course of the study the group of nurses remained unchanged.
In all, 361 patients were invited to participate in the study, of whom 298 agreed (83% response rate). Patients' progress through the study is presented in Figure 2. A nonrespondent analysis indicated that patients who declined to participate were significantly less well educated than those who agreed to participate (P< .05). During the study, 42 patients in the control group and 37 patients in the intervention group were lost to follow-up. The 2 most common reasons for dropout in both study arms were cessation of treatment and death. Data collection was completed within 13 months and 14 months, respectively, for the control group and the intervention group. Table 1 presents the baseline sociodemographic and clinical characteristics of the remaining 219 patients. These characteristics were well balanced between the 2 study arms.
|Control group (n=108)||Intervention group (n=111)||P|
|Mean age, y||55||57||.17|
|Low (<high school)||34||31|
|Middle (≥high school)||52||46|
|Mean COOP and WONCA scores (SD)†|
|Physical fitness||3.1 (0.9)||2.9 (0.9)||.1|
|Feelings||2.1 (1.1)||2.0 (1.0)||.67|
|Daily activities||2.5 (1.0)||2.4 (1.0)||.76|
|Social activities||2.1 (1.0)||2.1 (1.1)||.72|
|Overall health||3.2 (0.8)||3.2 (0.9)||.85|
|Pain||2.0 (0.9)||1.9 (0.9)||.43|
|Fatigue||2.4 (0.9)||2.5 (0.8)||.17|
|Quality of life||3.1 (0.8)||3.1 (0.8)||.86|
Communication about the patients' HRQL
On average, HRQL-related topics were discussed significantly more often in the intervention than in the control group. The mean (standard deviation [SD]) composite communication score at the fourth visit was 4.8 (3.3) in the intervention group and 3.8 (2.3) in the control group (P = .009). Table 2 reports the frequency with which specific HRQL topics assessed by the EORTC QLQ-C30, the BR23, and the CR38 questionnaires were discussed. The number of patients with lung cancer was too small to allow for a separate analysis of communication regarding topics addressed by the QLQ-LC13. Nine of the 13 HRQL issues assessed by the QLQ-C30, 5 of the 7 HRQL issues assessed by the BR23, and all 8 HRQL issues assessed by the CR38 were discussed more frequently during the fourth visit in the intervention group as compared with the control group. These differences reached statistically significant levels (controlling statistically for first study visit) for the discussion of dyspnea, sleep problems, constipation, diarrhea, cognitive functioning, breast-related complaints, and body image (breast cancer). Marginally statistically significant differences (P< .10) favoring the intervention group were observed for the discussion of colorectal cancer patients' gastrointestinal symptoms, urinary problems, and sexuality.
|QLQ-C30 Topics||Visit 1||Visit 4|
|Control group (n = 108)||Intervention group (n = 111)||P||Control group (n = 108)||Intervention group (n = 111)||P|
|QLQ-BR23 Topics||Visit 1||Visit 4|
|Control group (n = 65)||Intervention group (n = 52)||P||Control group (n = 65)||Intervention group (n = 52)||P|
|QLQ-CR38 Topics||Visit 1||Visit 4|
|Control group (n = 21)||Intervention group (n = 29)||P||Control group (n = 21)||Intervention group (n = 29)||P|
|Treatment-related side effects||24||35||.42||29||45||.55|
There were no statistically significant between-group differences in ICC scores at the first (baseline) visit (data not shown). At the fourth visit, nurses' awareness of the HRQL-related problems of the intervention group patients was significantly better (ie, significantly higher ICCs) than of the control group patients for daily activities (0.18 vs 0.44; P = .04), pain (0.14 vs 0.49; P< .001), and quality of life (0.24 vs 0.48; P = .05) (Table 3).
|Control group||Intervention group||P|
|Quality of life||0.24||105||0.48||107||.05|
HRQL-related medical chart notations and patient management activities
The mean total number of HRQL-related medical record notations (23 vs 20; P< .05), notations related to the QLQ-C30 functional scales (5.6 vs 4.1; P< .001), and notations related to the QLQ-CR38 scales (6.2 vs 3.3; P< .01) was significantly higher in the intervention group compared with the control group.
No statistically significant between-group differences were observed in the percentage of cases with HRQL-related medication prescription, test ordering, consultation with the treating oncologist, referrals to other healthcare providers, or modification/cessation of chemotherapy. A marginally significant difference favoring the intervention group was noted for HRQL-related advice/counseling (74% in the intervention group vs 62% in the control group; P = .06) (Table 4).
|Chart notations (mean)||Control group (n = 108)||Intervention group (n = 111)||P|
|QLQ-C30 functional scales||4.1||5.6||<.001|
|QLQ-C30 symptom scales||5.9||6.6||.16|
|EORTC QLQ Total‡||20.4||23.5||<.01|
|Action (% cases)|
|Ordering of tests||25||30||.43|
|Counseling or advice||62||74||.06|
No statistically significant group differences were observed for any of the patient satisfaction scales, with patients in both the intervention and control groups reporting high levels of satisfaction (data not shown).
There were no statistically significant between-group differences at the fourth visit on any of the SF-36 scales or on the FACT breast or colon cancer subscales (data not shown). There were insufficient numbers of lung cancer patients to perform statistical analysis of the FACT lung cancer scale.
Evaluation of the Intervention
Eighty-nine percent of the patients in the intervention group reported that the HRQL summary profile provided an accurate picture of their HRQL, 69% reported that it was used explicitly during treatment, 89% believed that the summary enhanced their nurses' awareness of their health problems, and 99% believed that it would be useful to introduce the intervention as a standard part of the outpatient clinic procedure.
All nurses reported that the summary provided useful information and indicated that it facilitated communication, especially with regard to ‘difficult subjects’ such as sexuality. They also reported that it resulted in a more efficient use of their time with the patients. All of the nurses indicated a desire to continue using the HRQL summary profiles in their daily practice.
As hypothesized, the routine collection and dissemination of HRQL-related information in the outpatient chemotherapy setting resulted in a significant increase in the frequency with which HRQL issues were discussed. The salutary effect of the intervention on patient-nurse communication is particularly encouraging in that not only observable symptoms were discussed more often, but also HRQL issues that are less observable (eg, sleep problems and cognitive functioning). The intervention also resulted in a (marginally) significant increase in the frequency with which several condition-specific HRQL issues were discussed (eg, body image in breast cancer; gastrointestinal and urinary symptoms, and sexuality in colon cancer).
Significant improvement was observed in nurses' awareness of the level of functioning (daily activities), symptoms (pain), and the overall quality of life of their patients. It is interesting to note that these are areas in which problems tend to be underestimated by healthcare providers.25, 26
The intervention also resulted in an increase in the number of medical chart notations related to HRQL issues, both generic and condition-specific. The intervention had only a modest effect on patient management activities, primarily in terms of increased levels of patient counseling. This could be due, at least in part, to the fact that no explicit practice guidelines were provided or recommendations made that linked (changes in) patients' self-reported HRQL with patient management strategies. Another possible explanation for the limited effect of the intervention on patient management is the relatively short duration of the intervention. Significant behavior change can require time, and thus longer exposure to the intervention, including the provision of feedback, might be expected to yield more positive results in this regard.27
The failure to detect significant between-group differences in patient satisfaction reflects, in large part, the high levels of satisfaction observed, which left little room for improvement. Such a ‘ceiling’ effect has been reported elsewhere.1, 28, 29 Satisfaction with the intervention itself was high, with most patients reporting that the HRQL summary profile was useful in facilitating communication and in enhancing nurses' awareness of patients' problems. Most patients and all of the nurses also favored introducing such a procedure as a standard part of the outpatient clinic procedure.
We are cognizant of several potential limitations of our study. First, although the patient sample was large, the nurse sample was relatively small. However, we have no reason to believe that the results are in any way atypical of what would be found with a larger group of nurses in that the participating nurses varied substantially in age and years of work experience.
Second, all things being equal, we recognize that a randomized design would have been preferred to the quasi-experimental design that we employed. However, for reasons described earlier, a randomized clinical trial was not feasible in the current clinical setting. It is important to note that the major potential threat to the internal validity of the study—a so-called ‘history effect’ because of changes in the nursing personnel or in the characteristics of patients treated within the hospital during the course of the study—did not materialize.
Third, the results pertaining to nurse-patient communication were based on self-report versus observational data. However, the self-report checklist that we used has been demonstrated to provide similar results to those based on observations of HRQL-related communication.
Finally, during the course of the study approximately one-quarter of the patients were lost to follow-up, primarily because of disease progression or death. However, this is not uncommon in studies of patients receiving palliative chemotherapy (making up approximately one-third of the patient sample).
In conclusion, the results of the current study support the use of standardized cancer-specific and site-specific HRQL assessments as a means of facilitating the discussion of HRQL issues, heightening nurses' awareness of their patients' problems, increasing the number of HRQL-related chart notations, and increasing counseling behavior. To increase the impact of this type of intervention on the more distal outcome of patients' HRQL, it may be necessary to identify critical thresholds for symptoms and functional impairment above which the use of patient management guidelines or clinical pathways would be triggered. This, in turn, requires highly precise HRQL assessments.30 Future studies aimed at combining HRQL assessments with practice guidelines should, for practical reasons, most likely focus on more homogeneous populations of patients (eg, women with early breast cancer). Otherwise, the identification of well-accepted practice guidelines for a broad range of cancer populations could prove difficult.
Improved health outcomes also depend on the ways in which patients' manage their own health. Additional interventions, such as the provision of tailored health education information based on patients' HRQL questionnaire responses and telephone coaching to promote self-management skills, are possible ways to enable patients to manage their own health concerns more effectively.31, 32
Finally, future studies are needed to investigate the possible use of routine HRQL assessment in the palliative care setting, in which symptom monitoring and management are of particular importance.
We thank Symone Detmar for input during the early stages of this research; Edith Budding, Carien de Molennaar, and Manon Komen for assisting in the data collection; and the patients, nurses, and oncologists of the oncology outpatient clinic of Medical Center Alkmaar for their willingness to participate in the study
- 14EORTC QLQ-C30 Scoring Manual. 2nd ed. Brussels, Belgium: European Organization for Research on the Treatment of Cancer; 1999., , , , .
- 32The clinical value of quality of life assessment in oncology practice — a qualitative study of patient and physician views. Psychooncology. 2007 Nov 21. [Epub ahead of print]., , , , , .