The role of nature in cancer patients' lives: a systematic review and qualitative meta-synthesis

The review focus was determined by the author’s doctoral thesis topic, which concerns the investigation of nature’s role in cancer patients’ lives and aims to determine nature-based care opportunities in cancer care contexts. The present paper reports research carried out by a sole researcher. The strategies employed to mitigate reporting bias are presented in the Limitations section below. For the purposes of this review, nature was defined as the phenomena of the physical world collectively, including various forms of vegetation and habitat, natural and humanly designed landscapes, natural cycles, processes, weather, wildlife and domestic animals, and other features and products of the earth including man-made creations which creatively organize and depict these nature elements [6, 26‐28].

The Cochrane Database for Systematic Reviews and the International prospective register of systematic reviews PROSPERO were first searched to ensure no identical or similar review was underway. The review was initially registered with PROSPERO as a systematic review of nature-based intervention research, however the literature searches returned insufficient studies addressing primary outcomes, which is a requirement for PROPERO registration. Consequently, the review focus shifted in order to address the available qualitative research, rather than intervention research. Registration (CRD42014015291) was withdrawn in February 2016.

A comprehensive systematic search of the published literature was conducted in following electronic database: Medline, CINAHL, PsycINFO and Cochrane Database of Systematic Reviews from January 1985 to March 2015. A research librarian was consulted for developing search strategies for the respective databases, which included combinatorial strings of Subject Headings and text word searches containing terms related to: “cancer” and “nature” (see Additional file 1 Search protocol). Electronic searches were supplemented by manual search of two relevant journals: AHTA Journal of Therapeutic Horticulture and ACTAHORT. Non-peer-reviewed articles retrieved from the electronic searches were read for the purpose of searching reference lists. The inclusion criteria for publications were: full-text, peer-reviewed journal articles published in English language, which included primary qualitative data of empirical studies conducted with cancer populations.

Initial searches retrieved titles and abstracts only. Duplicates, obviously irrelevant studies, and studies that did not meet the above inclusion criteria were removed. The next stage of the selection process was directed by reading records at full-text level and identifying whether the studies reported the use of qualitative methods to explore the experiences of contact with nature from the perspective of individuals who had experienced cancer. The study’s specific qualitative method employed or the researchers’ philosophical positioning (e.g. phenomenology, grounded theory, ethnography) were not directive because their shared focus is understood to be the elucidation of meaning and processes of a given phenomenon from the perspective of the experiencer through interpretive means [29]. Therefore, studies based on a predominantly quantitative research design with a minor qualitative supplement were not included as their aim precluded relevant data and interpretation to explore participants’ personal experiences.

Studies were eligible if the sample included people who had directly experienced cancer; excluded were studies reporting only from the perspective of caregivers or healthcare professionals. Studies focusing on related topics such as post-occupancy evaluations of hospital gardens [30] and physical exercise research [31], not exploring nature experiences were excluded.

Research synthesists are responsible for appraising the quality of included studies in order to report transparently on their validity and generalizability. Studies with methodological weakness will negatively impact on the strength of the conclusions in a meta-synthesis. Methodological discussion exists questioning the adoption of prescriptive evaluation protocols based on quantitative ontology and epistemology for evaluating qualitative research [32, 33]. In response, instruments are being developed to better reflect appraisal criteria relevant to qualitative research such as confirmability, dependability, and credibility [34]. Examples are the Joanna Briggs Institute Qualitative Assessment and Review Instrument (JBI-QARI) [35] and the Critical Appraisal Skills Programme (CASP) [36]. However, the use of prescriptive criteria remains controversial [33]. Sandelowski and colleagues recommend not excluding papers based on quality checklists, but to focus instead on topical relevance and to employ broad evaluation criteria for the final quality reporting of included studies [29]. Taking this discussion into account, the present meta-synthesis prioritized the following initial question to gauge topical relevance: Does this publication contribute genuine qualitative data exploring cancer patients’ nature experiences? The Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines [37] were considered an appropriately broad framework for assessing the quality of included studies, and they were also used as a guide for reporting findings in this present meta-synthesis. The strength of the COREQ guidelines lies in outlining three general domains of methodological rigour applicable to various qualitative methods, as demonstrated in one successful example of meta-synthesis by Luker and colleagues [38].

Author SB recorded data on the following characteristics of included studies: reference details (year of publication, author, first author’s country of origin); discipline; sample size; participant characteristics; methodological approach; data collection method; and focus of the study. All data presented as ‘results’ or ‘findings’ were entered into data management software QSR International’s NVivo 10 for Mac [39]. This included primary data (e.g. participant quotes) as well as authors’ interpretations (e.g. thematic description). All unaltered textual material was read in order to gain a general understanding of the material before inductively coding data line-by-line with the aim to glean salient underlying concepts. In this process, for example, the statement ‘escape from the fear and worry associated with cancer, a place that was safe’ [40] was reduced and captured with the descriptive label (code) ‘safe refuge’. Next, codes were grouped into meaningful clusters from which prominent descriptive themes emerged, for example, the codes ‘losing the capacity for gardening is painful’ [41], and ‘losing bond with garden’ [40] were grouped as ‘losing connection’. This interpretive task reflects terminology and descriptions based on the researcher’s own understanding of the material and is presented in Additional file 2 Themes and illustrative quotes, which includes reference to raw data in order to enable the reader’s own appraisal of interpretative and conceptual congruity.

Using a constant comparison approach [42], common events and attributes of the studied phenomenon were identified and careful attention was paid to the frequency of recurrence across different studies, which strengthens emerging themes [24]. For example, the descriptive theme ‘connecting with something outside’ arose in nine studies [13, 27, 28, 40, 41, 43‐46], while ‘stimulating sensory experiences’ was found in only four [27, 28, 44, 46]. Once the entire dataset was scrutinized (primary data and authors’ interpretations) and no more themes could be teased out, the researcher returned to the included publications for a second narrative reading in order to confirm contextual relationships between the themes. This informed the final theoretical synthesis of findings into overarching, analytical themes, or ‘core themes’ [47]. These core themes were developed in relation to the meta-synthesis’ main objective, which gave rise to a hierarchical pattern of core and descriptive themes organized according to their theoretical depth and relevance.

The systematic search identified 2342 records, 149 of which remained after duplicates and obviously irrelevant records were removed (Fig. 1). A further 77 of these records were eliminated by asking the question: Does this contribute genuine qualitative data exploring cancer patients’ nature experiences? The 68 remaining articles were read in full to determine the sample characteristics, the type of qualitative method, and whether the focus was indeed to explore nature experiences. Seventeen articles were identified as having substantial relevance and were further appraised against COREQ quality criteria. One of these articles was omitted because its study design was primarily quantitative [48]. Four studies lacked sufficient primary data and methodological rigour to produce understandings grounded in participants’ perceptions [49‐52]. Finally, one study did not differentiate clearly between the views of cancer patients and health professionals [53]. Overview of reasons for exclusion is provided in Additional file 3 Excluded publications. The 11 remaining documents were accepted for synthesis comprising ten articles (nine separate studies) and one thesis.

Table 1 summarizes the characteristics of the included documents. Studies were published between 2000 and 2014 with the majority (8/11) published after 2005. Data were contributed by 240 cancer patients across the lifespan with varying diagnoses and included survivors and palliative patients. Sample size ranged from 3 to 88, with seven studies reporting a sample size <16. Studies were conducted in four countries: United States (US) (n = 4), Canada (n = 4), United Kingdom (UK) (n = 2), and Australia (n = 1). Two publications by Unruh [40, 41] reported on different aspects of one data set collected in a single study.

The study designs and methodological rigour of the included articles were variable in quality. Notably, all articles omitted information about the researchers’ reflexive practice. As mentioned above, it has been argued that checklists may be an inappropriate instrument for evaluating qualitative research; however, the COREQ guidelines were used in this review as a broad guide for identifying unacceptably deficient methodology and provided a structured approach for consistent and fair treatment of dissimilar studies. The governing authority, nonetheless, remained the study’s subject relevance and whether it contributed genuine qualitative data. For example, according to the COREQ 32-item checklist, Rowlands’ [11] study scored only 11/32, yet it was found to contribute valuable insight about hospitalized cancer patients’ perceptions of connecting with the outside world when viewing nature through a window.

Based on the COREQ appraisal framework, a general comment can be made about the included studies regarding their consistent methodological weakness in the first domain (researcher’s reflexivity). In total, all studies combined scored 25/88 in this domain. Only two studies [41, 44] explained audit trails and strategies for verifying data analysis (participant checking). Overall, studies were rated considerably low in all three domains of the COREQ, and only three [13, 28, 41] scored individually >21 out of 32 appraisal items. Scoring appears in Additional file 4 Quality assessment.

The studies described a spectrum of nature experiences including views to nature from within the hospital [11], contact with therapeutically designed and natural landscapes [28, 54], domestic gardening and structured gardening programs [40, 41, 43, 45, 46], retreats in natural environments [55], dragon-boating [44], and outdoor adventure programs [13]. The initial coding process provided a broad sense of the material and explicated salient concepts. It was notable that individual researchers deployed different interpretative language when dealing with similar ideas and concepts, for example: experiences alluding to relaxation in nature were reported by English and colleagues as, ‘nature appears to inspire feelings of calmness’ [54]; while Unruh and colleagues reported, ‘Worried minds were eased by thinking about the garden’ [46]; and yet another study expressed the concept of relaxation as, ‘providing reprieve from the everyday stresses’ [45]. Consequently, this meta-synthesis prioritised primary data (participants’ own words) where available in order to start from, and remain close to, original formulations during the coding process. However, all extracted findings including individual authors’ interpretations were considered. Grouping of codes into logical clusters generated 22 descriptive themes. The shared and interrelated meanings of these descriptive themes informed the development of seven core themes: connecting with what is valued; being elsewhere; exploration, inner and outer excursions; safe and home; symbolism, understanding and communicating differently; benefitting from old and new physical activities; and, enriching aesthetic experiences. The core themes connecting with what is valued and being elsewhere, seeing and feeling differently pervaded all studies and the remaining five core themes were identified consistently in at least seven documents. Table 2 presents the thematic findings with citation to source documents and is followed by text summarizing each core theme with reference to descriptive themes in italics.

The present synthesis found that cancer patients valued contact with nature and benefitted from opportunities to connect with nature. Engaging with nature eased the strain related to cancer diagnosis and treatment by taking on several supportive roles: facilitating valued connections, transporting away from the burden and threat of cancer, encouraging inner and outer explorations, offering safe refuge, providing metaphoric material for understanding life changes, motivating physical activity, and enriching cancer patients’ lives aesthetically.

Comprehensive cancer care services need to consider patients’ values and experiences. Qualitative research is designed to generate a deep and broad understanding of human experiences and processes. Meta-synthesis can improve the translation of qualitative research into practice through locating, condensing and appraising relevant findings for the medical readership and healthcare management who govern clinical practice, research, and policy [61].

Patients’ motivations to seek nature, and the effects of these interactions were highly personal. The idiosyncratic associations between type of engagement and outcomes suggest that the benefits derived from engaging with nature cannot be predetermined nor administered. Contrary to what an interventionist approach would suggest [62], there is no indication to utilize nature in a prescriptive manner. However, cancer patients consistently attributed importance to engaging with nature and derived, in various forms, benefit and meaning from these interactions. The findings give credence to validating and enabling cancer patients’ own resources by appraising aspects of their lives and histories from which they draw meaning, strength, and hope. A range of practical examples from the collected literature suggest pathways for patients to access nature experiences, both in cancer care and home environments. Views to nature and nature art are easily incorporated in the design of clinical settings if put on the agenda, which patients reportedly derive benefit from. Hospital courtyards and home gardens can encourage various degrees of physical activity, motivate fruit and vegetable consumption, and create opportunities for connecting socially. Organized outdoor trips and activities can powerfully connect patients on a peer level and motivate positive lifestyle behaviours. While many examples are unsurprisingly simple and commonplace, further research is required to identify, understand, and safely implement such additional care opportunities for their maximum benefit.

The findings relate to cancer patients in westernized countries and cannot be said to transfer to other clinical or geographic populations. Although this review aimed at comprehensive sampling, the dearth of relevant literature resulted in a very small sample size (n = 11 documents), with a correspondingly small combined participant sample (n = 240). Based on Sandelowski’s recommendation that large sample size (exceeding 10 studies) can compromise “deep analysis” and “threaten interpretive validity” in meta-synthesis [29], the found literature was deemed sufficient to proceed with the synthesis.

Many of the included studies have serious methodological limitations, which must be considered when assessing the credibility of this review’s findings. In particular, the absence of audit trails, negative case reporting and researcher reflexivity challenge confirmability and present the possibility of bias in the recruitment and research procedures. To increase the review’s credibility, focus was placed on raw data (participants’ words) in order to keep the interpretations closer to participants’ own experiences.

Lastly, this review lacked an inter-rater process (such as member checking or peer debriefing), which is recommended in qualitative research to aid interpretation bias in work undertaken by sole researchers [42, 63]. While single author quantitative meta-analyses and qualitative meta-syntheses are uncommon, neither are unprecedented, see for example Dijkers [64] and Hammell [23]. Strategies were employed to mitigate this shortcoming and increase credibility. Firstly, it is clearly stated in the Methods section that this study represents a sole researcher study preparing the reader to navigate potential sole author bias. The findings remain traceable to their source materials by providing a considerable amount of primary data (audit trail) for internal validation and to enable the reader’s own audit of interpretations. Furthermore, detailed documentation of the search procedures and quality appraisals is made available to reflect rigour and the inclusion of topically relevant studies.