Abstract of the dissertation entitled. An Evidence-Based Guideline of Using Touch Therapy for Pain Relief. in Adult Cancer Patients.

Abstract of the dissertation entitled An Evidence-Based Guideline of Using Touch Therapy for Pain Relief in Adult Cancer Patients Submitted by Kwan Leung Ying for the Degree of Master of Nursing at The University of Hong Kong in July 2014 I considered a sub-acute public hospital setting of the Hong Kong Hospital Authority (HA), where adult cancer patients requiring pain control were often transferred to the unit for palliative care and pain control. Around 70% of these adult cancer patients reported moderate to severe pain. Touch therapy is an energy-based complementary therapy to facilitate the clients general health and well-being, and thus it is appealing to treat pain in cancer patients. However, there was no systematic review focusing on the pain relief effect of touch therapy in adult cancer patients. Therefore, this dissertation aimed to conduct a comprehensive systematic synthesis on the current

evidence of the effectiveness of touch therapy for pain relief in adult cancer patient, and develop an evidence-based guideline for using touch therapy in my target setting. Also, the feasibility and transferability of touch therapy in the target setting were discussed, and plan for implementation and evaluation were developed. Four electronic databases: CINAHL Plus, Cochrane Library, PsycINFO and PubMed were searched for randomized controlled trials that assessed the efficacy of touch therapy in reducing pain in adult cancer patients. After screening the titles, abstracts and subsequently full-texts, five eligible studies were identified. All of them had methodological quality rated as 1+ by the Scottish Intercollegiate Guidelines Network checklist. Based on their results, I concluded that there was enough evidence supporting touch therapy can effectively reduce pain in adult cancer patients. In the sequel, an evidence-based guideline of nine Grade A recommendations on using touch therapy for pain relief in adult cancer patients was developed. Touch therapy for pain relief in adult cancer patients is transferrable to the local sub-acute HA hospital, because the target setting shared similar patient characteristics and philosophy of care with that considered in the identified studies. Regarding feasibility, nursing staff in the target hospital department will have autonomy to carry out touch therapy. Moreover, equipment and facilities required for implementing touch therapy will be available. However, the expenditure of the new intervention includes

extra manpower arrangement and training program. To promote nursing staff participation, departmental year-rounded budget for in-service training which is already available should be allocated to support staff training of this new intervention. Nevertheless, implementing touch therapy will have potential benefits to adult cancer patients and nursing staff which outweigh the potential cost. The estimated net cost benefit obtained after discounting the set-up and running costs of touch therapy in the first year of service was HK$2,797,573. The stakeholders include the project team members, the Chief of Service (COS), the General Manager of Nursing (GMN), the Department Operation Manager (DOM), the Ward Managers (WMs), and the Outpatient Clinic Manager, as well as the case doctors and nursing staff in the department. They will be introduced to the change in the first 4-month period. The next 6-month period will be the facilitation stage to guide the change. There will be 6 Registered Nurses in the department attending overseas touch therapy training from the 7 th to the 9 th month. A 2-month pilot study will then be held. The evaluation of pilot study and proposed modifications will be included in the final guideline evaluation. The outcomes are categorized into 3 aspects including patient outcomes, healthcare provider outcomes and system outcomes. Patient outcomes are the pain score by a 0-10 point s Visual Analogue Scale (VAS), vital signs and patients satisfaction level. Healthcare provider outcomes are the satisfaction level

of all the nursing staff. System outcomes are the reduction of inpatient and outpatient cost of adult cancer patients with pain, as well as the overall expenditure to implement touch therapy. The implementation of touch therapy will take one year to determine if the defined outcomes are achieved.

An Evidence-Based Guideline of Using Touch Therapy for Pain Relief in Adult Cancer Patients Submitted by Kwan Leung Ying Bachelor of Science (Honors) in Nursing A thesis submitted in partial fulfillment of the requirement for the Degree of Master of Nursing at The University of Hong Kong July 2014

Declaration I declare that this thesis represents my own work, except where due acknowledgement is made, and that it has not been previously included in a thesis, dissertation or report submitted to this University or to any other institution for a degree, diploma or other qualifications. Signed... KWAN LEUNG YING i

Acknowledgements First and foremost, I would like to express my sincere gratitude to my supervisor, Dr. Daniel Fong as well as my former supervisor, Ms. Maureen Chui for spending so much time in reading my work and giving me advice in writing the paper. Thank you so much for the continuous support of my master study and the dissertation. This thesis would not have been possible without your patience and guidance. I would also like to express my deepest thanks to the staff of School of Nursing at the University of Hong Kong for the assistance and teaching in the past two years. I would like to take this opportunity to give my warmest thanks to my classmates, Ada, Cherry, Elaine, Michelle, and Iris for their accompany through each anxious day before the deadline of handing papers and holding presentations. I would like to thanks my husband Dennis for his love, support and patience through the journey of this master degree. Without his support, I would not have the courage to complete the study. Last but not least, I would like to say thanks to my family for their love and support throughout my life. ii

Contents Declaration... i Acknowledgements...ii Contents...iii-vi Chapter 1 Introduction 1.1 Background...1 1.2 Affirming the needs.. 2 1.3 Objectives & significances...6 Chapter 2 Critical Appraisal 2.1 Search and appraisal strategies 2.1.1 Identification of studies. 8 2.1.2 Selection criteria 8 2.1.3 Data extraction...9 2.1.4 Appraisal strategy....10 2.2 Results 2.2.1 Search Results..10 2.2.2 Studies characteristics 2.2.2.1 Study types.... 11 2.2.2.2 Sample characteristics...11 iii

2.2.2.3 Study country and setting. 12 2.2.2.4 Interventions. 12 2.2.2.5 Intervention provider....13 2.2.2.6 Duration and frequency of therapy and follow-up time...14 2.2.3 Quality assessment of studies...14 2.3 Summary and Synthesis 2.3.1 Summary..........17 2.3.2 Synthesis..........23 Chapter 3 Translation and application 3.1 Implementation potential 3.1.1 Target population and setting......28 3.1.2 Transferability.....28 3.1.3 Feasibility........31 3.1.4 Cost-benefit ratio..... 35 3.2 Evidence-based guideline......38 Chapter 4 Implementation plan 4.1 Communication plan 4.1.1 Communication with stakeholders..39 4.1.2 Process of communicating the plan. 40 iv

4.2 Pilot study plan 4.2.1 Preparation of pilot study........43 4.2.2 Procedures, data collection and data evaluation.44 4.3 Evaluation plan 4.3.1 Nature and number of clients to be involved.. 45 4.3.2 Outcomes.......46 4.3.3 Timing and frequency to take measurements.....47 4.3.4 Data analysis.....48 4.3.5 Basis for an effective change of practice....48 References......88 Appendices Appendix 1: Details of searching process 50 Appendix 2: Levels of evidence and grades of recommendations...51 Appendix 3: Quality assessment of studies......52 Appendix 4: Visual Analogue Scale (VAS) of pain......72 Appendix 5: Estimated time spent, cost and saving of touch therapy.. 73 Appendix 6: Guideline of Using Touch Therapy for Pain Relief in Adult Cancer Patients.....75 v

Appendix 7: Project calendar to illustrate the time frame of preparation, pilot study and implementation plan...82 Appendix 8: Satisfaction survey for nurses in the Palliative Care Unit..83 Appendix 9: Assessment form for patient undergoing touch therapy...84 Appendix 10: Questionnaire for adult cancer patients attending touch therapy course for pain relief... 87 vi

CHAPTER 1 INTODUCTION 1.1 BACKGROUND There were over 12.4 million diagnosed cancer cases worldwide in 2008 according to the World Cancer Report released by International Agency for Research on Cancer (IARC) in 2008 (Boyle, & Levin, 2008). In Hong Kong, 26390 cases were diagnosed with malignancy in 2010 (Hospital Authority, 2010). Owing to advances in cancer treatment, the survival rate of cancer patients has been increased with prolonged life span. However, cancer patients experienced pain with wide variations according to the primary diagnosis and disease stages (van den Beuken-van Everdingen et al., 2007). According to the definition of the International Association for the Study of Pain (2012), pain is an unpleasant sensory and emotional experience associated with actual or potential tissue damage. Approximately 30% patients with newly diagnosed cancer, 50% to 70% patients receiving active anti-cancer therapy, and 60% to 80% patients with advanced cancer stage experienced acute and chronic pain (Cherny, 2006). It was estimated that cancer-related pain was poorly managed in 80% of patients, which may affect their physical, psychological, social function, and vocational choices (Aghabati, Mohammadi, & Pour Esmaiel, 2008; Breivik et al., 2009; Harrington et al., 2010). 1

The current cancer-related pain management in pharmacological aspect follows the three-step analgesic ladder as suggested by the World Health Organization (WHO) with the initiation of non-opioids, weak to strong opioids with adjuvants prescribed whenever necessary on the basis of a pain rating scale (Newton, Hickey, & Jackowski, 2008; Sze, 2010). In Hong Kong, cancer-related pain control relies mainly on pharmaceutical aspects. However, analgesics can produce adverse effects and do not always succeed in eliminating pain (Abramowicz, 1993). In a survey conducted by Ward et al. (1993), cancer patients reported fears relating to analgesic use including addiction (75%), side effects (61%), fear of injections (65%) and tolerance (55%). Pain control with interventional approach has gained increasing interest due to the potential of decreasing adverse effects and possibly decreasing cost (National Center for Complementary and Alternative Medicine [NCCAM], 2013). 1.2 AFFIRMING THE NEEDS I considered a sub-acute public hospital setting of the Hong Kong Hospital Authority (HA), where adult cancer patients requiring pain control were often transferred to the unit for palliative care and pain control. There were around 40% of patients with lung cancer or other cancer types with different stages as current diagnosis. As the in-hospital beds of oncology units were limited in the leading acute hospital in 2

the same cluster of HA, patients requiring cancer induced or cancer treatment related pain control were often transferred to this unit for palliative care. For cancer patients undergoing anti-cancer therapy like chemotherapy or radiotherapy and suffering from cancer treatment related pain, as well as patients with distant metastases such as bone metastases, management of pain would be of major importance for them. In my unit, pharmaceutical treatment was the mainstay of pain control for adult cancer patients. The doctors were recommended to follow the WHO analgesic ladder, prescribing analgesic from level 1 (non-opioids) to level 3 (strong opioids) (Hospital Authority, 2007; Sze, 2010). According to daily pain level review on a 0-10 scale, around 70% of these adult cancer patients reported moderate to severe pain. However, opioids for pain control can often induce a number of side effects including sedation, inability to concentrate, agitation and respiratory depression (Hong Kong Cancer Fund, 2012). In view of the side effects, particularly when there was a risk of respiratory compromise, the doctors were reluctant to prescribe opioids unless the patients and family opted for end-of-life comfort care at the end stage of disease. Approximately 50% of patients reported fair pain control with analgesic prescribed based on clinical observation. The length of hospital stay was prolonged when patients refused to discharge owing to failure in pain control. In a cancer survey conducted by Wills and Wootton (1999), 52% of Hong Kong 3

Chinese cancer patients reported fears of addiction relating to analgesic use. In my unit, the choices of procedural interventions in pain control are mainly provided by the physiotherapists, which include transcutaneous electrical nerve stimulation (TENS), acupuncture, heat and cold treatment, and so on. A nurse initiated non-pharmacological evidence-based protocol for pain relief in adult cancer patients was not yet established in my unit. In a recent local study conducted by Tse et al. (2012), cancer patients could be more positive in handling their pain situations and took initiatives in the use of nonpharmacological strategies with pain management education provided by nurses. However, as nurses did not have the authority in analgesic dosage adjustment, we simply administered medication as doctors prescribed and referred patients to allied health services for interventional pain control. Apart from pain assessment, nurses role in pain control was minimal. Touch therapy is a completely non-invasive and non-pharmacological procedure to facilitate the clients general health and well-being, and thus it is appealing to treat pain in cancer patients. In a healthy individual, energy can flow evenly as the energy field is symmetrical and balanced (Jackson et al., 2008). When the energy field is imbalanced, physical and psychological symptoms can be arisen such as pain (Anderson, & Taylor, 2012). Touch therapy is an energy-based complementary therapy described by the National Center for Complementary and Alternative Medicine 4

(NCCAM, 2013). Therapeutic Touch (TT), Healing Touch (HT), and Reiki are the representing touch therapy under the category of complementary therapy of NCCAM (Coakley, & Barrone, 2012; So, Jiang, & Qin 2008). It is a potential comfort measure which supports allopathic, patient-centered, and holistic paradigms in health care and facilitate healing (Cook, Guerrerio, & Slater, 2004; Danhauer et al., 2008). In the United States, touch therapy was already identified as a core intervention for oncology nursing (Bulechek et al., 2012). Touch therapy practitioners use their hands on or above the clients body to direct healing energy, which in turns facilitate the clients general health and well-being through modification of energy field (Maville, Bowen, & Benham, 2008). Over the past two decades, over a hundred studies have reported positive outcomes and rarely recorded any side effects (Wardell, 2004; Winstead-Fry, 1999). Take Therapeutic Touch (TT) as an example, it is practiced by nurses in more than 70 countries and taught in 90 health facilities and 88 nursing and health-related schools (Krieger, 1993; Krieger, 1997). Nevertheless, touch therapy had not been well recognized and implemented in large scale health care institutions locally. The efficacy of touch therapy has been assessed in a systematic review but the review assessed publications up to 2008 only. However, it was not focusing on cancer patients and more subsequent evidence has emerged (So, Jiang, & Qin, 2008). Therefore, further search of relevant studies is needed and should be focused on the use of mainstream touch 5

therapy, i.e. HT, Reiki and TT, for pain relief in adult cancer patients. 1.3 OBJECTIVES AND SIGNIFICANCE The objectives of this dissertation were: 1. To conduct a systematic comprehensive synthesis on the current evidence of the effectiveness of touch therapy for pain relief in adult cancer patient; 2. To assess the implementation potential of touch therapy; 3. To develop an evidence-based guideline of touch therapy; and 4. To develop plan for implementing and evaluating the evidence-based guideline. Touch therapy, if proved to be effective, can be promoted as a nurse-initiated cancerrelated pain relief intervention. When nurses are familiarized with touch therapy, they can offer informed recommendations to cancer patients experiencing pain. It is therefore expected to contribute to my unit by shortening patients length of hospital stay, as well as minimizing expenditure with low cost equipment and facilities. Moreover, touch therapy can create professionalism in nurses role by introducing procedural interventions in cancer-related pain control. In this way, a trusting relationship between nurses and clients can be achieved. For patients, the use of analgesic can be reduced and medication-related side effects can be minimized. With 6

touch therapy offered as an extra option for pain control, patients satisfaction level can be enhanced and their comfort can be promoted during hospital stay. 7

CHAPTER 2 CRITICAL APPRAISAL 2.1 SEARCH AND APPRAISAL STRATEGIES 2.1.1 Identification of studies Four electronic databases: CINAHL Plus, Cochrane Library, PsycINFO and PubMed were searched for relevant randomized controlled trials (RCTs). The details of the search terms and search result are presented in Appendix 1. The search was narrowed down to RCTs only. Then, the titles were screened for relevancy, and the abstracts of potential studies were carefully read and examined. Afterwards, the full papers of potential studies were retrieved in order to confirm their eligibility. Furthermore, manual search of reference lists from those eligible studies was performed to avoid missing out any relevant studies. 2.1.2 Selection criteria The inclusion criteria for searching evidence were: 1. RCTs; 2. Studies involved the mainstream of touch therapy, i.e. Healing Touch, Reiki, and Therapeutic Touch; 8

3. Studies aimed at relieving cancer induced pain or cancer treatment related pain by touch therapy; 4. Studies that evaluated touch therapy on subjects pain or other related symptoms; 5. Studies involved subjects aged 18 or above regardless of the gender and types of cancer diagnosed. Studies not focusing on adult cancer patients were excluded. We used four groups of search keywords. The first group included touch therapy or biofield therapy or energy therapy or Healing Touch or Therapeutic Touch or Reiki. The second group included cancer or malignant/ malignancy or metastases. The third group included anti-cancer therapy or radiation therapy/ radiotherapy/ radiation treatment or chemotherapy. The forth group included pain or symptoms. There were no restrictions on language and publication year to avoid selection bias during the searching process. 2.1.3 Data extraction After reading the selected studies carefully, data was extracted with citation, study type, evidence level, sample characteristics, intervention, comparison(s), length of follow-up, outcome measures, and effect size included. The data was formulated as table of evidence (Table 1) following the alphabetical orders of the first author. 9

2.1.4 Appraisal strategy Methodology Checklist 2: Controlled Trials developed by the Scottish Intercollegiate Guidelines Network [SIGN] (2013a) was used as the appraisal tool for quality assessment of each RCT, providing that RCT design was one of the selection criteria (Appendix 3). Studies would be rated as high quality (++) if most criteria were met with little or no risk of bias. Studies would be rated as acceptable (+) if most criteria were met with some flaws and an associated risk of bias. Studies would be rejected if either most criteria were not met or significant flaws relating to key aspects of study design was noted. In additions, the levels of evidence (ranged from 1++ to 4) and the corresponding grade of recommendation (ranged from A to D) would be rated according to SIGN grading system (1999-2012) (Appendix 2) (SIGN, 2013b). The levels of evidence of each study are also shown in the table of evidence (Table 1). 2.2 RESULTS 2.2.1 Search Results The comprehensive systemic search was performed from 8 th August to 13 th September in 2013. Total 312 citations were extracted from the databases for further review. After title screening, 21 studies were identified as potentially relevant. The abstracts of these 21 potential studies were carefully examined. Also, all of their full 10

papers were retrieved to screen for eligibility. At last, five eligible studies were identified with population, interventions and outcomes corresponded to my study. Their reference lists were read but revealed no additional relevant studies. Table 1 summarizes the five studies. 2.2.2 Studies characteristics 2.2.2.1 Study types All of the five studies were RCTs involving the randomization of subjects, comparing between two groups (i.e. intervention versus placebo or control) (Cook, Guerrerio, & Slater, 2004; Olson, Hanson, & Michaud, 2003; Pohl et al., 2007), among three groups (Aghabati, Mohammadi, & Esmaiel, 2010) and four groups (Post-White et al., 2003). One study performed within-subject comparison with crossover design (Post-White et al., 2003). 2.2.2.2 Sample characteristics The mean ages of patients ranged from 41 (Aghabati, Mohammadi, & Esmaiel, 2010) to 57.3 (Olson, Hanson, & Michaud, 2003). The median age was 60 in one study (Pohl et al., 2007). 10.9% to 37% of patients were diagnosed with stage I to II cancer, and 21% to 49% of patients were diagnosed with stage III to IV cancer (Aghabati, Mohammadi, & Esmaiel, 2010; Cook, Guerrerio, & Slater, 2004; Post-White et al., 11

2003). All studies recruited patients with metastatic cancer or stage IV cancer. Four of the studies recruited subjects who were undergoing anti-cancer therapy (Aghabati, Mohammadi, & Esmaiel, 2010; Cook, Guerrerio, & Slater, 2004; Pohl et al., 2006; Post- White et al., 2003), and one study recruited subjects for palliative symptom management (Olson, Hanson, & Michaud, 2003). 2.2.2.3 Study country and setting The studies were carried out from four countries including Austria (Pohl et al., 2007), Canada (Olson, Hanson, & Michaud, 2003), Iran (Aghabati, Mohammadi, & Esmaiel, 2010), and the United States (Cook, Guerrerio, & Slater, 2004; Post-White et al., 2003). The settings of studies included in-hospital medical unit (Pohl et al., 2007), inpatient hospice, oncology or palliative unit (Aghabati, Mohammadi, & Esmaiel, 2010; Olson, Hanson, & Michaud, 2003); outpatient clinic (Olson, Hanson, & Michaud, 2003; Post-White et al., 2003), and hospital radiotherapy center (Cook, Guerrerio, & Slater, 2004). 2.2.2.4 Interventions All the five studies evaluated the effect of touch therapy, including Healing Touch (HT) (Cook, Guerrerio, & Slater, 2004; Post-White et al., 2003), Reiki (Olson, Hanson, & Michaud, 2003), Therapeutic Touch (TT) (Aghabati, Mohammadi, & Esmaiel, 2010), and laying on of hands (Pohl et al., 2007) on pain relief in adult cancer patients. For 12

HT, the procedures included healer preparation, pre-treatment energetic assessment, HT interventions (unruffling, magnetic unruffling, full-body connection, mind clearing, chelation, and lymphatic drain to modulate the energy field), and post-treatment energetic assessment (Cook, Guerrerio, & Slater, 2004; Post-White et al., 2003). For Reiki, the traditional Usui method was used which was delivered to 18 specific areas of the body (Olson, Hanson, & Michaud, 2003). For TT, the procedures included centering, assessment, TT administration which comprised of directing, modulating, change pattern in human energy, and reassessment (Aghabati, Mohammadi, & Esmaiel, 2010). For laying on of hands, the therapy provider performed energy-based touch therapy similar to TT as claimed by the researcher (Pohl et al., 2007). 2.2.2.5 Intervention provider Touch therapies were performed by therapy providers with different training levels among five studies, including 45 hours certified training (Cook, Guerrerio, & Slater, 2004), 5 months of training by touch therapy specialist (Aghabati, Mohammadi, & Esmaiel, 2010), practitioner credential program (Post-White et al., 2003), and master level program (Olson, Hanson, & Michaud, 2003). There was a self-declared healer who performed the intervention without the qualification of touch therapy mentioned (Pohl et al., 2007). Registered Nurses were acted as therapy provider in one study (Post- White et al., 2003). 13

2.2.2.6 Duration and frequency of therapy and follow-up time The durations of intervention ranged from 5 minutes (Pohl et al., 2007), 30 minutes (Aghabati, Mohammadi, & Esmaiel, 2010; Cook, Guerrerio, & Slater, 2004), 45 minutes (Post-White et al., 2003) to 90 minutes (Olson, Hanson, & Michaud, 2003). The frequency of therapy administration ranged from consecutive days (Aghabati, Mohammadi, & Esmaiel, 2010), alternate days (Pohl et al., 2006) to two days apart (Olson, Hanson, & Michaud, 2003). The length of follow-up ranged from 5 days (Aghabati, Mohammadi, & Esmaiel, 2010) to 56 days (Cook, Guerrerio, & Slater, 2004; Post-White et al., 2003). 2.2.3 Quality assessment of Studies In general, all the five studies addressed an appropriate and clearly focused question with purposes and/or objectives covered. Hypothesis was clearly stated in four studies (Aghabati, Mohammadi, & Esmaiel, 2010; Cook, Guerrerio, & Slater, 2004; Olson, Hanson, & Michaud, 2003; Post-White et al., 2003). The assignments of subjects were randomized to intervention and comparison(s) by different methods, including card numbers selected from different pockets (Aghabati, Mohammadi, & Esmaiel, 2010), computer-generated random number assignment system (Olson, Hanson, & Michaud, 2003), computer-assisted program and predefined stratification criteria (Pohl 14

et al., 2007), and table of random number (Cook, Guerrerio, & Slater, 2004). One study addressed that randomization was performed but it did not mention the method of randomization (Post-White et al., 2003). Owing to the explicit nature of touch therapy, blinding was not mentioned or performed in most of the studies, except in one study which blinded study subjects by comparing touch therapy with a placebo touch procedure and used an opaque screen to separate treatment providers (Cook, Guerrerio, & Slater, 2004). However, the study did not ascertain itself as investigator-blinded, as the study coordinator might inadvertently figure out the group assignment. Another study was originally designed to be subjectblinded but the therapy provider violated the agreement and unblinded the information to the subjects (Pohl et al., 2007). One study mentioned that research nurse, participants, and project biostatistician who perform data analysis did not know the group assignment (Olson, Hanson, & Michaud, 2003). However, it also mentioned that it was not possible to blind given the explicit nature of intervention. Therefore, it was uncertain whether blinding was successful in the remaining four studies. In additions, all the five studies did not address the use of concealment method. The demographic characteristics for the intervention and comparison(s) were statistically comparable in four studies (Aghabati, Mohammadi, & Esmaiel, 2010; Cook, Guerrerio, & Slater, 2004; Olson, Hanson, & Michaud, 2003; Post-White et al., 15

2003). There were no subjects dropping out in one study (Aghabati, Mohammadi, & Esmaiel, 2010). Three studies had 20.5% to 29% of subjects dropped out (Cook, Guerrerio, & Slater, 2004; Pohl et al., 2006; Post-White et al., 2003). The highest dropout rate 54.7% was noted in one study, as the research team considered subjects with Mini-mental state examination (MMSE) below 23/30 as ineligible and they were withdrawn from the study (Olson, Hanson, & Michaud, 2003). The dropout reasons among five studies included increasing severities of illness (Cook, Guerrerio, & Slater, 2004; Olson, Hanson, & Michaud, 2003; Pohl et al., 2006), complications (Pohl et al., 2007), family problems (Cook, Guerrerio, & Slater, 2004), time arrangement failure (Cook, Guerrerio, & Slater, 2004; Post-White et al., 2003), religious objections (Cook, Guerrerio, & Slater, 2004), lacking of compliance (Pohl et al., 2007), preferred different treatment or no longer meeting the crossover criteria (Post-White et al., 2003). With end-stage cancer patients included, discontinuation due to death was reported in three studies (Olson, Hanson, & Michaud, 2003; Pohl et al., 2006; Post-White et al., 2003). The intention to treat analysis was performed in two studies, with mean substitution as the single imputation method for missing data point or including random block factors (Cook, Guerrerio, & Slater, 2004; Pohl et al., 2006). After quality assessment was performed according to Methodology Checklist 2: Controlled Trials, all studies met most criteria of RCTs and acceptable for further 16

review (SIGN, 2013a). All of them had methodological quality rated as 1+ by the SIGN checklist (Appendix 2). The characteristics of studies were summarized in the following parts. 2.3 SUMMARY AND SYNTHESIS 2.3.1 Summary All the five studies investigated the use of touch therapy for pain relief in adult cancer patients. Bone metastasis was the main source of pain (38%) from one study which only recruited those patients with pain score 3 out of 10 (Olson, Hanson, & Michaud, 2003). Scale of 0 to 10 was used to measure pain for the cancer patients, including Visual Analogue Scale (VAS) of pain (Aghabati, Mohammadi, & Esmaiel, 2010; Olson, Hanson, & Michaud, 2003; Pohl et al., 2006), and Brief Pain Index (BPI) (Post-White et al., 2003). One study used SF-36 for Health-related quality of life with pain as subscale (Cook, Guerrerio, & Slater, 2004). The overall effect for pain relief among five studies was statistically significant. All studies showed statistically significant pain relief effect when comparing intervention to control, and intervention to placebo. One study showed significant pain relief effect over 5 days when comparing intervention to placebo scored by 10cmVAS pain scale (p=0.04, n=90) (Table 1) (Aghabati, Mohammadi, & Esmaiel, 2010). In 17

another study, 16.1% reduction in pain score was noted when comparing intervention to placebo scored by SF-36 for Health-related quality of life (p<0.02) (Table 1) (Cook, Guerrerio, & Slater, 2004). Three studies compared the pain score between intervention group and control group scored by 10-point VAS scale or BPI, with intervention groups producing 5.5% to 15% pain relief effect higher than control group (p<0.03) (Table 1) (Olson, Hanson, & Michaud, 2003; Pohl et al., 2006; Post-White et al., 2003). One study tested the immediate pain relief effect over all touch therapy sessions computing by area under the curve (AUC) scored by BPI with significant result obtained (p<0.0001) (Table 1) (Post-White et al., 2003). Two studies adopted vital signs as objective measurements (Olson, Hanson, & Michaud, 2003; Post-White et al., 2003). When comparing intervention to control, one study had statistically significant reduction in diastolic blood pressure with 5mmHg difference (p=0.005) and mean heart rate with 7 bpm difference (p=0.019) (Table 1) (Olson, Hanson, & Michaud, 2003). Another study showed statistically significant reduction of heart rate (-6 to 7.2bpm), respiratory rate (-1.7 to 2.8rpm), systolic (-8 to 11mmHg), and diastolic blood pressure (-3 to 4.5mmHg) when comparing intervention to control over four days (p<0.001) (Table 1) (Post-White et al., 2003). 18

Table 1. Table of evidence Citation/ Study Type/ Evidence Level 1. Aghabati et al., 2010 -RCT 10 ; pre-test/post-test design; (1+) Sample Characteristics Intervention Comparison(s) Length of Followup Outcome measures -Mean age: 40.95 -TT 13 (n=30): Centering, 1.Control: Routine 5 days (1) Pain, scored by (n=90, 0 dropout) Assessment, TT 13 administration interventions in ward 10cmVAS 14 pain scale -100% female (directing, modulating, & change (n=30) (2) Fatigue, scored by -All were undergoing pattern in human energy), 2. Placebo: Mimic TT 13 10cm RFS 11 fatigue chemotherapy at Reassessment (n=30) scale inpatient oncology units -Room with privacy provided -30 minutes per 5 days - All with GCS 5 =15 -Lying position -Same movements as -Cancer Stage: -Performed by researcher with 5 TT 13 by simply began and II: 18% months training counted back from 100 III: 33% -30 minutes per 5 days IV: 49% -Duration was determined after expert consultation with TT 13 specialist and the clinical expertise with cancer patients and considering samples psychological and physical status. Effect size Intervention versus comparison(s) Intervention vs placebo: (1) Days: 1 st, 2 nd, 3 rd, 4 th, 5 th (p<0.04) (2) Days: 1 st, 2 nd, 3 rd, 4 th, 5 th (p<0.03) Intervention vs control: (1) Days: 1 st, 2 nd, 3 rd, 4 th, 5 th (p<0.0001) (2) Days: 1 st, 2 nd, 3 rd, 4 th, 5 th (p<0.001) 19

2. Cook et al., -Mean age: 50.7±13.1 -HT 6 (n=44, 10 dropout): -Placebo: Mimic HT 6 56 days (1) Pain Intervention vs placebo: 2004 (n=78, 16 dropout) Healer preparation (e.g. centering, (n=34,6 dropout) (2) Vitality (1) 70.6±23.6, 54.5±29 -RCT 10 ; -100% female attuning), pretreatment energetic - Lay person of HT 6, did (3) Physical role (p<0.02) Single-blinded (to - All were newly assessment (e.g. hand scan), not extend hands over the functioning (2) 54.6±21.4, 41.9±22 subjects); pre-test/ diagnosed gynecological HT 6 intervention, post-treatment body of any subject Scored by SF-36 for (p<0.03) post-test design or breast cancer energetic assessment -Not to focus thoughts on Health-related quality of (3) 70.9±24.7, 57.7±27.6 (1+) -All were undergoing no -Lying position & fully clothed subject life (p<0.05) more than 1/3 radiation -Performed by Level II providers -Conduct simple therapy at hospital with 45-hours training mathematical calculations radiotherapy center -Combined with usual medical and in their heads -Cancer Stage: nursing services -Combined with usual I: 35% -Immediately after radiation medical and nursing II: 37% treatment services III or IV: 21% -6 sessions (30 minutes each) -6 sessions (30 minutes each) 3. Olson et al., -Mean age: 57.3 -Standard opioid plus Reiki (n=11 -Standard opioid plus 7 days (1) Pain, scored by 10- Intervention vs control: 2003 (n=53; 29 dropout) excluded dropout) rest (Control) (n=13 point VAS 14 pain score (1) Day 1 st : -1.2, -0.3 -RCT 10 ; -% gender not provided - Traditional Usui method was used excluded dropout) with ESS 4 Staging tool (p=0.035) pre-test/ post-test -All with MMSE 7 23 -Delivered to 18 specific areas of -Rest for 90 minutes after (2) Psychological Day 4 th :-1.5, -0.1 design -All VAS 14 for pain 3 the body. Ten hand positions were first afternoon analgesic component, scored by (p=0.002) (1+) -Source of pain: performed on the head and torso. dose 10-point VAS 14 score (2) Day 1 st to 7 th : +0.8, 0 Bone metastasis: 38% - Eight additional hand positions, for QoL 9 Assessment (p=0.002) Neuropathic pain: 8% covering the back, hip area, and feet (3) Mean diastolic blood (3) Day 1 st : -4, +1 (p=0.005) Lymphadenopathy: 8% -Lying position pressure (mmhg) (4) Day 1 st : -7, 0 (p=0.019) Ascites: 4% -Performed by Reiki Master (4) Mean heart rate Visceral: 4% (bpm 2 ); 20

Unknown: 38% -90-minutes Reiki on Day 1 & 4 - All were cancer after first afternoon analgesic dose patients from inpatient palliative unit (primary site), hospice, and outpatient symptom management clinic 4. Pohl et al., -Median age: 60 -Laying on of hands by a self- -Mimic procedure by an 10 days (1) Pain Intervention vs control: 2007 (n= 80; 23 dropout) declared healer (n=40 excluded actor (Control) (2) Depression (1) -0.67 (p=0.03) -RCT 10 ; Two- -65% female dropout) (n=40 excluded dropout) (3) Relaxation (2) -1.07 (p=0.02) group -All were cancer patients -Techniques similar to TT 13-5-minutes duration (4) Inner peace (3) -1.33 (p=0.04) experimental pre- from in-hospital medical -1 st session on Day 1, 2 nd session on -3 days per week (5) General wellbeing (4) -0.67 (p=0.02) test/ post-test unit with ECOG 3 Day3, 3 rd session on Day 5 Scored by 10-point VAS 14 (5) -0.53 (p=0.002) design performing status 2 (Secondary endpoint), with 5-days Well-Being scale with (1+) -Metastases: treatment free interval until Day 10 subscales Brain: 5% (primary endpoint) Bone: 21.3% -Patients were treated by healers at Visceral: 72.5% 1 st & 2 nd sessions, and treated by -All were receiving anti- actors since the 3 rd session because cancer therapy (chemo-, the healer quitted unexpectedly. radio-, hormone or - 5-minutes duration immunotherapy) 5. Post-White et -Mean age:54.7 -HT 6 : Centering, unruffling, 1.MT 8 : Swedish massage 56 days (1) Mean heart rate Intervention vs MT 8, presence al., 2003 -n=230 (66 dropout) magnetic unruffling, full-body protocol with defined (bpm 2 ) & control: -RCT 10 ; -86.1% female connection, mind clearing, strokes followed with Day 1 st, 2 nd, 3 rd, 4 th 21

pre-test/post-test; -All were cancer patients chelation, and lymphatic drain to massage gel applied. (2) Mean respiratory rate (1) -6.4, -6.4, -6, -7.2 crossover design receiving chemotherapy modulate the energy field. -4 weekly 45-minutes (rpm 12 ) (p<0.001) (1+) with an identical -Pre-intervention symptom and sessions (3) Mean systolic blood (2) -1.7, -1.8, -2.8, -2.8 repeating cycle for 2 vital signs were assessed. (n=63) pressure (mmhg) (p<0.001) remaining cycles -Room with privacy & soft music 2.Presence: (4) Mean diastolic blood (3) -8, -10, -10, -11 (p<0.001) -All were from outpatient provided. -Patients lie for 45 minutes pressure (mmhg) (4) -3, -4, -4.5, -3 (p<0.001) chemotherapy clinic -Performed by certified & on the same table used for (5) Different of mean (5) -0.75, -0.8, -0.55, -0.75 -Cancer Stage: credentialed practitioner who were the MT and HT 6 sessions. scores, scored by (p<0.011) I: 10.9% Registered Nurses. (n=45) Brief Pain Index (BPI) (6) p<0.0001 II: 23.9% -4 weekly sessions, 45 minutes each 3.Control: Standard (6) Immediate effect over (7) p=0.003 III:14.8% (n=56) cancer treatment 4 sessions tested by (8) p=0.028 IV: 43.9% (n=164) AUC 1, scored by BPI Unstaged: 6.5% (7) Total mood -Pain rating 3 over 10: disturbance, scored by 71% the Profile of Mood States (POMS) (8) Fatigue, scored by POMS Abbreviations: AUC 1 : Area under the curve; bpm 2 : Beat per minute; ECOG 3 : Eastern Cooperative Oncology Group; ESS 4 : the Edmonton Staging System; GCS 5 : Glasgow Coma Scale; HT 6 : Healing Touch; MMSE 7 : Mini mental state examination; MT 8 : Therapeutic Massage; QoL 9 : Quality of Life; RCT 10 : Randomized Controlled Trial; RFS 11 :Rhoten Fatigue Scale; rpm 12 : rate per minute; TT 13 : Therapeutic Touch; VAS 14 : Visual Analogue Scale 22

2.3.2 Synthesis Among the five studies, the study populations represented a spectrum of adult cancer patients with different cancer stages. All studies recruited patients with metastatic cancer or stage IV cancer. Four of them recruited subjects who were undergoing anti-cancer therapy (Aghabati, Mohammadi, & Esmaiel, 2010; Cook, Guerrerio, & Slater, 2004; Pohl et al., 2006; Post-White et al., 2003), and one study recruited subjects for palliative symptom management (Olson, Hanson, & Michaud, 2003). Through integrated review, the five studies were of moderate methodological quality. The effect of touch therapy in adult cancer patients showed statistically significant pain relief effect among all studies. The outcomes are favorable to the target population in the palliative care unit who are seeking for cancer induced or cancer treatment related pain relief. Since nurses in Hong Kong do not have the authority in prescribing medication, touch therapy can be used as an adjunct treatment with standard opioid prescribed by doctors if indicated as one study proposed (Olson, Hanson, & Michaud, 2003). Another study also provided evidence that touch therapy can be combined with usual medical and nursing services (Cook, Guerrerio, & Slater, 2004). None of the reviewed studies highlighted the need of doctor-initiated in performing touch therapy. Besides, Registered Nurses were acted as therapy provider in one study with evidence to support touch therapy as a nurse-led intervention in the proposed 23

guideline (Post-White et al., 2003). Given that touch therapy was already identified as a core intervention for oncology nursing in western countries (i.e. the United States), the proposed clinical practice guideline should be performed by nurses after touch therapy training provided (Bulechek et al., 2012). Among the five studies, the training level of the therapy providers varied from 45 hours to 5 months of certified training (Cook, Guerrerio, & Slater, 2004) to master level (Olson, Hanson, & Michaud, 2003). 45 hours training hours can be set as the minimum requirement for nurses as a qualified practitioner to perform touch therapy in my unit. As touch therapy is well established in western countries, overseas training with international recognized qualification will be promoted to interested nursing staff after considering manpower arrangement. According to the two studies investigating pain relief effect of Healing Touch by Cook, Guerrerio, & Slater (2004) and Post-White et al. (2003), both of them investigated subjects undergoing anti-cancer therapy or with metastatic cancer, which were the target population in my unit. As touch therapy is a new intervention in my unit, training is an essential element for consideration. In view of the manpower arrangement and a relatively short-term training duration (i.e. 45 hours) to obtain the international recognized qualification, Healing Touch will be selected for implementation in my unit. Besides, considering evidence supported that Healing Touch can be performed with 24

fully clothed allowed, it can reduce unnecessary exposure of the clients body and clients hesitancy to the procedures. Touch therapy session will begin with healer preparation and pre-treatment energetic assessment. Afterwards, the therapy provider will perform the energetic techniques which include centering, unruffling, magnetic unruffling, full-body connection, mind clearing, chelation, and lymphatic drain to modulate the energy field of the clients (Cook, Guerrerio, & Slater, 2004; Post-White et al., 2003). Upon the completion of the therapy session, a post-treatment energetic assessment will be performed. To ensure privacy during the therapy session, environmental preparation should also be considered. Privacy with room or screen provided can prevent interruption during the procedure (Post-White et al., 2003). In order to promote comfort, patients can undergo therapy in lying position as three studies suggested (Aghabati, Mohammadi, & Esmaiel, 2010; Cook, Guerrerio, & Slater, 2004; Post-White et al., 2003). Pre-intervention and post-intervention pain level assessments were performed by Visual Analogue Scale (VAS) of pain scale of 0 to 10, which was used by three out of five selected studies (Aghabati, Mohammadi, & Esmaiel, 2010; Olson, Hanson, & Michaud, 2003; Pohl et al., 2006). The three studies mentioned that the pain scale was well developed and validated, with reliability statistically comparable with previous 25

studies. The 0-10 scale is clear and simple for patients to evaluate pain level. Also, it was currently used in my unit for daily pain assessment. Therefore, it can be used as a consistent and standard pain measurement scale for nursing staff. Vital signs (i.e. blood pressure, pulse, respiratory rate) will be used as objective indicators which will be assessed before and after intervention as suggested by two studies (Olson, Hanson, & Michaud, 2003; Post-White et al., 2003). To enhance the pain relief effect, touch therapy can be performed immediately after cancer treatment (Cook, Guerrerio, & Slater, 2004) or after analgesic is administered (Olson, Hanson, & Michaud, 2003). The duration of intervention as provided by the five studies ranged from 5 minutes to 90 minutes. Considering cancer patients psychological and physical status, a session with intervention around 30 minutes will be provided. Besides, 30 minutes should be considered as a reasonable duration of intervention allowing more therapy sessions per day to meet the demand of patients. The target population included patients with endstage cancer who may have rapid deterioration in health condition and increased in pain level. Considering the relatively high dropout rate related to the poor disease prognosis and fair treatment compliance of the reviewed studies, a shorter follow-up period is recommended to facilitate the compliance of outcome evaluation of target population. Since the proposed follow-up period is short, the frequency of therapy sessions can be set as consecutive days with regard to clients availability and pain level, as well as 26

considering manpower arrangement. In conclusion, the reviewed studies revealed sufficient evidence to show touch therapy as a nurse-led pain relief intervention can benefit adult cancer patients in current hospital setting. Therapy selection, training level, environmental considerations, pain level measurement, as well as the duration, frequency and follow-up period of touch therapy were synthesized from all the reviewed studies and carefully selected for implementation. 27

CHAPTER 3 TRANSLATION AND APPLICATION 3.1 IMPLEMENTATION POTENTIAL 3.1.1 Target population and setting The systematic review in Chapter Two identified adequate evidence supporting that touch therapy is effective for pain relief in adult cancer patients. The new intervention is potentially transferrable to the local sub-acute HA hospital, because the target setting shares similar patient characteristics and philosophy of care with that considered in the identified studies. The target population of touch therapy will be adult cancer patients who are aged 18 or above, and suffering from cancer induced pain or cancer treatment related pain. The department at the local setting consists of two male wards, two female wards and an outpatient clinic. Touch therapy sessions will be provided to the cancer patients who required pain relief in both inpatient ward settings and the outpatient clinic. 3.1.2 Transferability Records of patients admitted at the local hospital department between December 2012 and December 2013 were manually retrieved from the clinical management 28

system at the target setting. The age range of cancer patients in the local setting was 45 to 71 years. The age range of subjects in the five reviewed studies was 41 to 60 years, which is comparable to those in the local target population. With regard to gender, the female proportion was 65% to 100% in the reviewed studies. In the target setting, the ratio of male patients to female patients is one to one. There are equal numbers of gender specific wards, and similar proportion of patients with both genders visit the outpatient clinic. There were two out of five reviewed studies recruiting 100% proportion of female patients. One study recruited female patients because of the disease nature (i.e. gynecological or breast cancer) (Cook, Guerrerio, & Slater, 2004). Another study was conducted in country with Moslems cultural belief. It recruited female patients because there were only female practitioners available to serve female clients (Aghabati, Mohammadi, & Esmaiel, 2010). In the target setting, the types of cancer include gynecological cancer and other nongender-specific cancer types like lung cancer. Patients with all types of cancer are indicated to undergo touch therapy for pain relief purpose. In fact, 70% of cancer patients in the target unit have moderate to severe pain level and the majority of them have metastatic cancer, which meet the sample characteristic of the reviewed studies and may show the transferability of the intervention. In the target setting, the ratio of male nursing staff to female nursing staff is about 3 to 7. Although both male and female staff can offer touch therapy to patients 29

after training, male nursing staff will not perform touch therapy on female patients to avoid potential allegations of sexual harassment. In the identified studies, the pain rating of subjects were 3 in a 10-point scale (Olson, Hanson, & Michaud, 2003; Post-White et al., 2003). The metastatic cancer type was dominant in four out of five studies which can induce severe pain (Aghabati, Mohammadi, & Esmaiel, 2010; Olson, Hanson, & Michaud, 2003; Pohl et al., 2006; Post-White et al., 2003). In the target setting, 70% cancer patients reported pain as moderate to severe level (i.e. 5-8 out of 10), which was similar to those subjects in the identified studies. However, as all the studies were performed in western society, cultural difference exists since the target population will be mainly Chinese in Hong Kong. As an interventional approach therapy, touch therapy is non-invasive in nature and patients can simply rest in bed to receive the nursing procedure with simple instructions to follow. Therefore, the impact of cultural diversity is minimal and the new intervention can be applied in Hong Kong society without obvious obstacles. The philosophy of care is considered as a major determinant of transferability of the new evidence-based intervention. With regard to the vision and mission of Hospital Authority (2013a), the staff was committed to the promotion of the health of community by empowering them to regain control over their health as well as eliminating their length of stay in the hospital. By offering support with professional services, the staff 30

should take the initiatives to serve patients in a better way by improving skills and stay abreast with the latest health care developments. Since the vision and mission of the target hospital and department is to serve and to provide high quality care to patients, implementation of touch therapy can go along with the value of Hospital Authority and the target hospital and department where nurses are trained for touch therapy as a new intervention for pain relief in adult cancer patients. The number of patients whom can obtain benefits from the intervention should be considered for the transferability of the new intervention. As official data were not released by the target hospital and department, estimation was based on clinical observation and manual checking of the patient admission records in the four wards and outpatient attendance records. There were 700 patients diagnosed of cancer, with approximately 500 out of them have the need of pain relief. It is estimated that 500 patients can gain the benefit from the proposed intervention. 3.1.3 Feasibility Nursing staff in the target hospital department will have autonomy to carry out touch therapy. The nurse managers like the General Manager of Nursing (GMN), the Department Operation Manager (DOM), the four Ward Managers (WMs) and the Outpatient Clinic Manager in the department will take the major lead for approving, 31