Level of agreement between patient self-report and observer ratings of health-related quality of life communication in oncology

https://doi.org/10.1016/j.pec.2006.06.002Get rights and content

Abstract

Objective

To determine the level of agreement between patients and observers regarding the frequency with which health-related quality of life topics are discussed during outpatient clinical oncology visits.

Methods

The sample (n = 50) consisted of a consecutive series of cancer patients undergoing chemotherapy. Both the patients and observers used a checklist to report which HRQL topics had been discussed during the consultation. Percentage of agreement, kappa and adjust-kappa statistics were calculated.

Results

The percentage agreement between patients’ and observers’ ratings was generally high, ranging from 74% for fatigue to 96% for sleep problems and cognitive functioning. The average percentage of agreement over the 13 HRQL topics rated was 86%. Cohen's kappa varied between 0.41 (for pain) and 0.78 (for sleep problems). Prevalence-adjusted kappa's were generally higher, ranging from 0.48 (for fatigue) to 0.92 (for sleep problems and social functioning). The average Cohen's kappa and prevalence-adjusted kappa over the 13 HRQL topics were 0.56 and 0.71, respectively. Level of agreement was not found to vary significantly as a function of patients’ background characteristics.

Conclusion

Oncology patients’ self-reports of the HRQL-related topics discussed during outpatient chemotherapy visits are in reasonably close agreement with those provided by observers.

Practice implications

Our results indicate that the patient is a legitimate source of information about the HRQL-related content of medical encounters, and thus can be used in communication studies where the collection of observational data (e.g., via audio- or videotaping) is either too costly or logistically impractical.

Introduction

The most optimal way to evaluate health care communication at the individual level is to make use of audiotapes or videotapes [1]. However, audiotaping or videotaping conversations may not always be feasible for either technical or privacy reasons. For example, when treatment is given in open oncology wards where more than one person is treated at the same time, it may be difficult to isolate the relevant conversations to be recorded and there may be concerns with privacy and confidentiality. In addition, recording and evaluating tapes can be costly and time-consuming.

Alternative methods used for evaluating communication include medical chart reviews and patient self-report. Although medical record audits may be very useful in documenting observable behaviors linked to communication (e.g., medication prescription and referrals), they are less helpful in evaluating more general communication behavior, such as inquiries and advice about symptoms and functional limitations, as these are typically not noted systematically in the medical record [2], [3].

Patient self-report data are frequently used to evaluate communication [4], [5], [6]. Collecting self-report data is relatively efficient and low cost. However, such self-report data also have their limitations. Patients’ recall of conversations can be misinformed by lapses in memory and may be distorted when sensitive topics are at issue. In addition, patients’ worries and concerns may limit their ability to understand and to remember details of their medical visit [1]. For these reasons, although they are widely used, the validity of patient self-report measures of health care communication can be questioned when used as the primary or only source of such information. Only a few studies have directly compared patients’ self-reports with audio- or videotapes of their medical visits (for a general review, see Boon and Stewart [7]). Several studies report that patients tend to overestimate the occurrence of their physicians’ counseling regarding alcohol and tobacco habits [8], [9], [10]. Dimatteo et al. have reported that patients overestimate physicians’ counseling behavior, but underestimate the discussion of medications [2].

Over the past decade, communicative exchanges between health care providers and patients have become an area of increasing interest for research in oncology [11], [12]. An essential condition for optimal cancer treatment is that physicians, nurses and other ancillary health care providers are able to communicate effectively with their patients in order to obtain as complete a picture as possible of the patients’ physical and psychosocial health status.

Health-related quality of life (HRQL) considerations have come to play an increasingly important role in selecting among available treatment options and in monitoring the effects of treatments over time [13], [14], [15]. HRQL is typically defined in terms of physical, psychological and social functioning, and symptom experience [13]. Yet, recent studies have reported that HRQL issues are often not addressed during outpatient medical visits in oncology. For example, Detmar and colleagues reported that, during outpatient palliative treatment consultations, physicians devoted 64% of their conversation to medical/technical issues and only 24% to HRQL issues. Importantly, in 20–54% of consultations in which patients were experiencing serious HRQL problems, no time at all was devoted to the discussion of these problems. In particular, patients’ fatigue and emotional functioning often went unaddressed (in 48% and 54% of the cases, respectively) [16].

A number of studies have investigated methods for improving HRQL-related communication. These efforts have focused primarily on training health care providers in communication skills, both as a part of the medical and nursing school curriculum, and in continuing education programs [17], [18], [19], [20]. More recently, interventions have been developed that provide health care providers with individualized, standardized, patient self-reported HRQL information as a means of facilitating communication in daily clinical practice [21], [22]. In order to evaluate the effect of these efforts, the actual patterns and content of communication between patients and health care providers have to be evaluated in a systematic way.

To the best of our knowledge no studies have assessed the extent of agreement between patient self-reported and observational data on HRQL-related communications. The current study was undertaken to determine whether, in studies of patient-health care provider communication, it is possible (i.e., valid) to rely on patients’ self-reports of what actually was discussed during medical consultations. This is important because there are situations where it may not be possible to actually observe (or record) such conversations, and thus one may need to rely on the patients’ self-report. To investigate the validity of patient self-report data on HRQL-related communication, comparisons were made with data obtained via observer-raters. Given that patient characteristics may influence self-reports, we also examined whether the level of agreement between the self-reported and observed communication was associated with various sociodemographic and clinical characteristics, including age, education, sex, and treatment intent.

Section snippets

Study site and subjects

The study was conducted in the outpatient chemotherapy clinic of the Department of Internal Medicine of the Medical Center Alkmaar, a large community hospital located north of Amsterdam. In this clinic patients receive chemotherapy either in a private room or in the ward where a maximum of six patients are treated at the same time. Oncology nurses are responsible for the delivery of treatment, and the day-to-day management of patients.

The patient sample was composed of a consecutive series of

Sample characteristics

A total of 57 patients were invited to participate in the study, of whom 50 agreed (response rate = 88%). Of the seven non-participants, four declined due to very poor physical condition, one reported insufficient interest or lack of time, and two had other reasons not specified.

The demographic and clinical characteristics of the patient sample are shown in Table 1. Forty-six percent was female, with a mean age of 59 years (range = 27–79 years). Patients had a wide range of cancer diagnoses, with

Discussion

This study was undertaken to investigate the level of agreement between patients and observers with regard to the frequency with which HRQL-related topics are discussed in daily clinical oncology practice. An additional objective was to determine whether the level of agreement between patients and observers varied systematically as a function of the sociodemographic and clinical characteristics of the patients.

As has been suggested in the literature [16], acute and more observable symptoms

Acknowledgements

The authors wish to thank Saskia de Bie, Edith Budding and Carien de Molennaar for assisting in the data collection. We thank the patients and the nursing staff of the oncology outpatient clinic of the Medical Center Alkmaar for their willingness to participate in this study.

References (29)

  • P. Taenzer et al.

    Impact of computerized quality of life screening on physician behaviour and patient satisfaction in lung cancer outpatients

    Psychooncology

    (2000)
  • R.L. Street et al.

    Using health status surveys in medical consultations

    Med Care

    (1994)
  • J. Wasson et al.

    The short-term effect of patient health status assessment in a health maintenance organization

    Qual Life Res

    (1992)
  • A. Adams et al.

    Alcohol counseling: physicians will do it

    J Gen Intern Med

    (1998)
  • Cited by (11)

    • Communication about Sexual Health in Breast Cancer: What Can We Learn from Patients’ Self-Report and Clinic Dialogue?

      2020, Patient Education and Counseling
      Citation Excerpt :

      Self-report measures are easy to administer and score and are cost-effective, but may be subject to potential recall errors and offer less rich data than time-intensive measures (e.g., observation). In contrast with other areas of quality of life [23], to our knowledge, no studies have examined how well aligned patients’ self-report and observed clinic dialogue are for clinical communication about sexual health. We surmised that an analysis into discrepancies between patients’ self-report and clinic dialogue might offer insight into how patients interpret clinical discussions of sexual health.

    • Different perspectives on communication quality and emotional functioning during routine oncology consultations

      2012, Patient Education and Counseling
      Citation Excerpt :

      In oncology, it has been shown that different perspectives (patient, physician, and observer) have different views of what had been discussed in relation to phase I clinical trial discussions [32]. Furthermore, oncology patients’ self-reports of health-related quality of life-related topics discussed during outpatient visits have been shown to be in reasonably close agreement with those provided by observer [33]. More research comparing methods in relation to each other (method triangulation) has been suggested for oncology care [33].

    • What do symptom scores mean: Observations on discrepancies when defining symptoms using words and numbers

      2010, European Journal of Oncology Nursing
      Citation Excerpt :

      Statistically, neither result showed a significant difference in preference. Patient self-report is considered the gold standard for evaluating patients’ symptoms, functional capacity and quality of life (Hilarius et al., 2007). Understanding and acting on these responses, however, assumes that HCPs interpret responses the way patients intended, and that patients have accurately represented their symptoms.

    • Facing others in pain: Why context matters

      2018, Social and Interpersonal Dynamics in Pain: We Don't Suffer Alone
    View all citing articles on Scopus
    View full text