The Effectiveness of Shared Care in Cancer Survivors—A Systematic Review

Objectives: To determine whether the shared care model during the follow-up of cancer survivors is effective in terms of patient-reported outcomes, clinical outcomes, and continuity of care. Methods: Using systematic review methods, studies were searched from six electronic databases—MEDLINE (n = 474), British Nursing Index (n = 320), CINAHL (n = 437), Cochrane Library (n = 370), HMIC (n = 77), and Social Care Online (n = 210). The review considered all health-related outcomes that evaluated the effectiveness of shared care for cancer survivors. Results: Eight randomised controlled trials and three descriptive papers were identified. The results showed the likelihood of similar effectiveness between shared care and usual care in terms of quality of life, mental health outcomes, unmet needs, and clinical outcomes in cancer survivorship. The reviewed studies indicated that shared care overall is highly acceptable to cancer survivors and primary care practitioners, and shared care might be cheaper than usual care. Conclusions: The results from this review suggest that the patient satisfaction of shared care is higher than usual care, and the effectiveness of shared care is similar to usual care in cancer survivorship. Interventions that formally involve primary care and improve the communication between primary care and hospital care could support the PCPs in the follow-up.

British Nursing Index search strategy (ab(follow up) OR ab(follow-up) OR ab(followup*) OR ab(postsurgery) OR ab(post-surgery) OR ab(postsurgical*) OR ab(post surgical*) OR ab(postoperat*) OR ab(post-operat*) OR ab(continuity of patient care) OR ab(disease management) OR ab(surveillance) OR ab(routine test*) OR ab(disease progression) OR ab(aftercare) OR ab(survivorship)) AND (ab(shared care) OR ab(sharing of care) OR ab(co-management) OR ab(collaborative care) OR ab(care coordination) OR ab(coordinated care) OR ab(referral AND consultation) OR ab(cooperative behavio*r) OR ab(delivery of health care) OR ab(shared service*)) AND ((SU.EXACT("Cancer : Pain") OR SU.EXACT("Lung Cancer") OR SU.EXACT("Colorectal Cancer") OR SU.EXACT("Cervical Cancer") OR SU.EXACT("Cancer : Services") OR SU.EXACT("Cancer : Children") OR SU.EXACT("Cancer : Nursing") OR SU.EXACT("Cancer : Radiotherapy") OR SU.EXACT("Skin Cancer") OR SU.EXACT("Cancer : Chemotherapy") OR SU.EXACT("Testicular Cancer") OR SU.EXACT("Head and Neck Cancer") OR SU.EXACT("Cancer : Surgery") OR SU.EXACT("Cancer : Counselling") OR SU.EXACT("Cancer") OR SU.EXACT("Ovarian Cancer") OR SU.EXACT("Breast Cancer") OR SU.EXACT("Prostate Cancer") OR SU.EXACT("Cancer : Alternative Therapies")) OR cancer* OR neoplas* OR malignan* OR carcinoma* OR sarcoma* OR oncolog* OR tumo*r* OR adenocarcinoma* OR infiltrat* OR medullary OR intraductal) (MM "Shared Services, Health Care") OR "shared care" 686 S33 S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20   How was this carried out, some methods may produce broken allocation concealment • Was the allocation concealed from researchers? Score: 2 The randomization was conducted by an independent professional centre after both patients and their GPs had signed informed consent, so the allocation was concealed from researchers. Although as mentioned in the paper, there might be potential selection bias because the relationship between patients and practitioners could affect the patients" decision about participation. However, the author also provide a reasonable explanation so that the rigor of the study is not affected. 3. Were patients, health workers and study personnel blinded? Consider: • Health workers could be; clinicians, nurses etc • Study personnelespecially outcome assessors Score: 0 As for this research, it is impossible to blind patients and practitioners because they are aware of the whole process. But the study did not mentioned about keeping outcome assessors blinded, which is a shortage of this research. 4. Were the groups similar at the start of the trial? Consider: Look at • Other factors that might affect the outcome such as age, sex, social class, these may be called baseline characteristics Score: 2 The important baseline factors such as age, stage of cancer, and treatment type had been considered and well balanced. The author also stratified the allocation in the randomization since patients came from different treatment centres. 5. Aside from the experimental intervention, were the groups treated equally? Score: 1 This study had considered about balancing two arms. All patients received five follow-up visits, the difference between two group was two specialist follow-up replaced by the GPs. The only difference was the GPs visited the patients at the beginning to re-engage with them, which might affect the results. 6. Were all of the patients who entered the trial properly accounted for at its conclusion? Consider: • Was the trial stopped early? • Were patients analysed in the groups to which they were Score: 2 No evidence showed the trial stopped early. All patients with results had been analysed properly. Score: 2 All outcome measurements were patient-reported questionnaires. The reliability and validity had been discussed in the research protocol. The primary outcomes included four questionnaire, a single specific question about patients" preference of care model, and a financial assessment. All results had been reported clearly, and no significant difference between two groups expect the single specific question (P<0.001). Besides, no evidence indicated selective reporting. 8. How precise was the estimate of the treatment effect? Consider:

Appendice 5 Example of using CASP RCT checklist to appraise a selected RCT
• What are the confidence limits? • Were they statistically significant Score: 2 The 95% confidence interval were provided in the study, and the sample size could provide 80% power to detect differences of 0.6 SD at two side. The significant differences were defined as P<0.05. Was the need for this trial clearly described?
Score: 1 Most important information has been reported, and the needs also clearly described. The limit was the number of patients changed in the outcome report table, but the author did not explain the reason. Besides, the article did not mentioned how to deal with incomplete questionnaires. 11. Are the benefits worth the harms and costs? Consider: • Even if this is not addressed by the trial, • what do you think?
Score: 2 Since the results in this research showed that shared care could provide similar outcomes to usual care with lower costs, the benefit worth all the efforts.

Appendice 6 Example of using HCPRDU quantitative research checklists to appraise a selected quantitative study
Bibliographic Details: Lund, Shared Care in prostate cancer: a three-year follow-up, SCANDINAVIAN JOURNAL OF UROLOGY, Denmark, 2016 [35] Review Area Score and Reason (1) STUDY OVERVIEW Score: 1 This is a multicentre research which lasted for three years. The bibliographic details were provided, and the key finding was clearly defined. However, the weaknesses are the definition of "shared care" was not detailed and the objective of this study was not clear.
(2) STUDY, SETTING, SAMPLE AND ETHICS Score: 1 The limitation is the research type was not mentioned in the content, and there was no comparison intervention. Besides, whether the sample size was sufficient has not been defined, which could not warrant the conclusions drawn. The study was based on the shared care model, and the patient discharge summery proceeded as the standard of follow-up recommendation, which was provided to the GPs. The inclusion and exclusion criteria were described. The patients were enrolled from three hospitals and they were stratified according to hospital, age, and treatment, the dropout was also distributed to three hospitals, which could warrant the conclusion. Besides, all the outcomes were illustrated clearly as well as how to deal with the missing data.
(3) ETHICS Score: 1 The paper did not mention about the ethical approval, but all patients had given their consent. (4) DATA COLLECTION, ANALYSIS AND POTENTIAL RESEARCHER BIAS Score: 1 The author had considered about the confounding variables when analysing the data. The outcome measures were appropriate in the study. All data was collected and provided the sufficient details about the meaning. Besides, the length of follow-up was three years, which was sufficient to detect the effects. Although the non-responders could lead to some bias, the dropout was equally distributed in terms of hospital and age, and there was no difference between non-responders and participants. The limit is that there was no comparison and the validity and reliability of questionnaires were not clarified, which might affect the credibility of the conclusion. (5) POLICY AND PRACTICE IMPLICATIONS Score: 2 The study findings underlined shared care could increase the patient and GP compliance in cancer follow up, which could be widely utilised in practice. (6) OTHER COMMENTS Score: 2 The references were listed in the paper. In generally, it provided a feasible way of follow up for cancer patients based on a multicentre and long term research. Total Score: 8 GP=general practices "0" represents many limitations, "1" represents some limitation, "2" represents excellent. The aim and key findings were described in the text. The strengths of the study is that the intervention was innovative and clearly defined, which provide a new way to carry out shared care between hospital nurses and community nurses. The weakness was that the quantitative outcome measurements and results were not clearly specified, and no data was reported.
(2) STUDY AND CONTEXT (SETTING, SAMPLE AND OUTCOME MEASUREMENT) Score: 1 This study is mixed type study and the intervention was quite clear and the design is very creative. There is adequate detail in intervention group and comparison group. The rationale for the study programme was explained and the shared care was home-cased care majorly conducted by community nurses. As for sample, the author did not mentioned about how many participants were involved in the quantitative procedure. Besides, the outcome measurement was not mentioned in the text.
(3) ETHICS Score: 1 The study obtained ethical approval and oral informed consent was provided to the patients who joined the qualitative interview. But whether the participants signed informed consent was not mentioned. (4) GROUP COMPARABILITY Score: 0 The author mentioned the data form three distinct time points would be collected and compared to hospital activity data, but no relevant data could be found in the paper. (5) QUALITATIVE DATA COLLECTION AND ANALYSIS Score: 1 The data collection methods were provided, but the process of fieldwork such as the interview guide and length was not adequately described. The description of data analysis process was mentioned but not sufficient. The findings were quite clear. The interview and the coding process were conducted by the researcher team, so there might be some potential bias. (6) POLICY AND PRACTICE IMPLICATIONS Score: 1 The cancer survivors are increasing in the whole world. The nurses play an important role in the management of cancer patients. This research provide a good way to combine oncology nurses and community nurses in the integrated care. However, the programme might cost a lot of money, which is the major obstacle of implementation. (7) OTHER COMMENTS Score: 1 Although the design of the study was quite good in this mixed study, but more details about the data collection and data analysis in both quantitative and qualitative part should be covered.