Background

People with cancer experience a variety of symptoms as a result of their disease and the therapies involved in its management. Inadequate symptom management has implications for patient outcomes including functioning, psychological well-being and quality of life (QoL). Attempts to reduce the incidence and severity of cancer symptoms have involved the development and testing of psycho-educational interventions to enhance patients' symptom self-management. With the trend for care to be provided nearer patients' homes, telephone-delivered psycho-educational interventions have evolved to provide support for the management of a range of cancer symptoms. Early indications suggest that these can reduce symptom severity and distress through enhanced symptom self-management.

Objectives

To assess the effectiveness of telephone-delivered interventions for reducing symptoms associated with cancer and its treatment. To determine which symptoms are most responsive to telephone interventions. To determine whether certain configurations (active ingredients, dosage) of telephone interventions mediate observed cancer symptom outcome effects.

Search methods

We searched the following databases: the Cochrane Central Register of Controlled Trials (CENTRAL; 2019, Issue 1) in the Cochrane Library; MEDLINE via OVID (1946 to January 2019); Embase via OVID (1980 to January 2019); CINAHL via Athens (1982 to January 2019); British Nursing Index (1984 to January 2019); and PsychINFO (1989 to January 2019). Additionally, we searched conference proceedings to identify published abstracts, and SIGLE and trial registers for unpublished studies. We also searched the reference lists of all included articles for additional relevant studies. Finally, we hand searched the following journals: Cancer, Journal of Clinical Oncology, Psycho-oncology, Cancer Practice, Cancer Nursing, Oncology Nursing Forum, Journal of Pain and Symptom Management, Palliative Medicine. Search was restricted to publications published in English.

Selection criteria

We included randomised control trials (RCTs) and quasi-RCTs which compared one or more telephone interventions with each other, or with other types of interventions (e.g. a face-to-face intervention) and/or usual care, with the stated aim of addressing any physical or psychological symptoms of cancer and its treatment, which recruited adult (over 18 years) men and women with a clinical diagnosis of cancer, regardless of tumour type, stage of cancer, type of treatment and time of recruitment (e.g. pre, during or post treatment).

Data collection and analysis

Two review authors independently selected articles, extracted data, and appraised methodological quality and risk of bias. Disagreements were resolved through discussion, involving the entire review team where necessary. Risk of bias was assessed using the Cochrane's risk of bias tool. We had planned to conduct meta-analyses using random effects models for symptoms where there were sufficient data to enable this. Heterogeneity between study outcomes was planned to be determined through visual inspection of forest plots and calculation of the I2 statistic. Where possible, outcomes are reported as standardised mean differences (SMDs) with 95% confidence intervals (CIs) and a descriptive synthesis of study findings is presented. Findings are reported on according to symptom addressed and intervention type (e.g. telephone only or telephone combined with other elements). As many studies had small samples, and baseline scores for study outcomes often varied for intervention and control groups, change scores and associated standard deviations were used.

Main results

Thirty-two studies were eligible for the review; most had moderate risk of bias, mostly related to blinding. Collectively they recruited 6250 people. Interventions were studied in people with a variety of cancer types and across the disease trajectory although many participants had breast cancer, early stage cancers and/or were starting treatment. Symptoms measured were anxiety, depression, emotional distress, uncertainty, fatigue, pain, sexuality-related symptoms and general symptom intensity and/or distress.

Interventions were primarily (n = 24) delivered by nurses, most (n=16) had a background in oncology, research or psychiatry. Ten were delivered solely by telephone; the rest combined telephone with additional elements (face-to-face consultation, and digital/online/printed resources). Number of calls delivered ranged from 1 to 18; most provided 3 to 4 calls.

Twenty‐one studies provided evidence on effectiveness of telephone‐delivered interventions and the majority appeared to reduce symptoms of depression compared to control. Nine studies contributed quantitative change scores (CSs) and associated standard deviation results (or these could be calculated). Likewise, many telephone interventions appeared effective when compared to control in reducing anxiety (16 studies; 5 contributed quantitative CS results); fatigue (9 studies; 6 contributed to quantitative CS results); and emotional distress (7 studies; 5 contributed quantitative CS results). Due to significant clinical heterogeneity with regards to interventions introduced, study participants recruited, and outcomes measured, meta‐analysis was not conducted.

For other symptoms (uncertainty, pain, sexuality-related symptoms, dyspnoea and general symptom experience) there was limited evidence; meta-analysis was similarly not possible and the results from individual studies were largely conflicting making conclusions about their management through telephone-delivered intervention hard to draw. There was considerable heterogeneity across all trials for all outcomes.

Overall, the certainty of the evidence was low for all outcomes in the review due to uncertainty over estimates. Outcomes were all downgraded due to concerns about overall risk of bias profiles being frequently unclear and due to inconsistencies in results and general heterogeneity.

Unsubstantiated evidence suggests that telephone interventions in some capacity may have a place in symptom management for adults with cancer. However, in the absence of reliable and homogeneous evidence, caution is needed in interpreting the narrative synthesis. Further, there were no clear patterns across studies regarding which forms of interventions (telephone alone versus augmented with other elements) are most effective. It is impossible to conclude with any certainty which forms of telephone intervention are most effective in managing the range of cancer‐related symptoms that people with cancer experience.

Authors' conclusions

Telephone interventions provide a convenient way of supporting self-management of cancer-related symptoms in adults with cancer. They are becoming more important with the shift of care closer to patients' homes, need for resource/cost containment and potential for voluntary-sector providers to deliver healthcare interventions. There is some evidence supporting use of telephone-delivered interventions for symptom management in adults with cancer; most evidence relates to four commonly experienced symptoms - depression, anxiety, emotional distress and fatigue. Some telephone-delivered interventions were augmented by combining them with face-to-face meetings and provision of printed or digital materials. The review was unable to determine whether telephone alone or in combination with other elements provides optimal reduction in symptoms; it appears most likely that this will vary by symptom. It is noteworthy that, despite the potential for telephone interventions to deliver cost savings, none of the studies reviewed included any form of health economic evaluation.

Further robust and adequately reported trials are needed across all cancer-related symptoms as the certainty of evidence generated in studies within this review was largely low to moderate, and reporting was of variable quality. Efforts are needed by researchers to reduce variability between studies in future. Studies in the review were characterised by both clinical and methodological diversity; the level of diversity hindered comparison across studies. At the very least, efforts should be made to standardise outcome measures. Finally, studies were compromised by having small samples, inadequate concealment of group allocation, lack of observer blinding and short length of follow up. Consequently, conclusions relating to symptoms most amenable to management by telephone-delivered intervention are tentative.