Diabetes Management in Patients Receiving Palliative Care

Vol. 32 No. 3 September 2006

Journal of Pain and Symptom Management

275

Original Article

Diabetes Management in Patients Receiving Palliative Care Karen Quinn, RN, MN, Peter Hudson, RN, PhD, and Trisha Dunning, AM, RN, MEd, PhD, FRCNA Center for Palliative Care (K.Q., P.H.); Department of Endocrinology and Diabetes (T.D.), St. Vincent’s Health, University of Melbourne, St. Vincent’s Health (K.Q., P.H., T.D.), Fitzroy, Victoria, Australia

Abstract A literature review revealed no evidence-based guidelines specific to managing diabetes in the context of palliative care. The purpose of the current project was to describe the management practices of doctors and nurses caring for people with diabetes and advanced disease. Palliative care doctors, palliative care nurses, endocrinologists, and diabetes nurse educators participated in this study. A two-phase project was undertaken: 1) two focus groups, and 2) a cross-sectional survey using a self-completed questionnaire. The focus group and questionnaire data identified that doctors and nurses used a range of practices and blood glucose testing frequencies to control blood glucose based on experience and not according to a robust evidence base. Implications for practice include the importance of collaboration between diabetes and palliative care specialists, and the need to develop clinical management guidelines. J Pain Symptom Manage 2006;32:275e286. Ó 2006 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Key Words Diabetes, palliative, health professionals, advanced cancer

Introduction Worldwide, 171 million people have diabetes and that number is expected to double by 2010.1 Ten million people worldwide were newly diagnosed with cancer in 2005.2 The

The Nurses Board of Victoria (NBV) provided the funding for this project. The views expressed in this paper are not necessarily those of the NBV. Address reprint requests to: Karen Quinn, RN, MN, Center for Palliative Care, University of Melbourne and Caritas Christi Hospice, Box 65, St. Vincent’s Hospital Melbourne, P.O. Box 2900, Fitzroy, Victoria 3065, Australia. E-mail: [email protected] Accepted for publication: March 20, 2006. Ó 2006 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved.

incidence of diabetes in patients with cancer is postulated to be six times that found in the general population.1 Medical advances in disease management have resulted in people living longer, often with multiple morbidities, eventually reaching a stage at which palliative management is paramount. Palliative care refers to the care of people with advanced life-threatening diseases that are no longer responsive to curative treatment, and the goal of care focuses on management of symptoms to achieve optimal quality of life for patients and their families.2 The causes of diabetes in patients with advanced disease are multifactorial and include the increased incidence of diabetes and advanced disease in older people, diabetogenic 0885-3924/06/$esee front matter doi:10.1016/j.jpainsymman.2006.03.011

276

Quinn et al.

medicines such as corticosteroids, obesity, and metabolic changes due to cancer.3e5 Patients with diabetes are not able to efficiently utilize glucose due to a deficiency or lack of insulin. Consequently, stores of fat and protein are mobilized to supply energy, and weight loss, weakness, and lethargy develop unless the insulin is enhanced or replaced to control blood glucose.4 Patients with advanced cancer have a reduced interest in food and are often fatigued. Both these symptoms make it difficult to achieve appropriate energy requirements and optimal glycemic control, which are essential for a patient with diabetes.4 Some medicines commonly used in advanced cancer, specifically corticosteroids and some diuretics, are known to trigger hyperglycemia.5 Several symptoms of advanced cancer, such as nausea and lethargy, also occur as a consequence of hyperglycemia, which makes it difficult for health professionals to determine the underlying cause of hyperglycemia. Careful monitoring and control of blood glucose levels via diet and glucose-lowering medicines (insulin and/or oral agents) are essential components of glycemic control.4 There are evidence-based clinical management guidelines and metabolic targets for patients with diabetes.4 However, as outlined in the following literature review, it appears that glycemic management in the context of advanced cancer may vary, and lacks a suitable evidence base.

Literature Review A literature search was conducted using the MEDLINE (1966e2005) and CINAHL (1982e2005) databases. The principal search terms used, either singly or in combination, were diabetes, hyperglycemia, hypoglycemia, end-of-life, palliative care, terminal illness, advanced cancer, hospice, corticosteroid, insulin, sliding scale, blood glucose, dexamethasone, and management. Articles were excluded from the literature review if they related to glucose homeostasis, pharmacological interactions, or were not in English. Palliative care and diabetes textbooks and their reference sections were also hand searched for relevant information. After reviewing all abstracts, 18 journal articles appeared to be relevant to the project. However, following full text review,


only four articles were directly relevant to managing blood glucose in palliative patients.5e8 Three of these four papers5e7 were based on expert opinion and clinical experience rather than research. McCoubrie et al.8 conducted a retrospective audit of 42 patient medical records before developing guidelines for managing diabetes in patients with advanced cancer. All patients included in the audit had a diagnosis of both diabetes and advanced cancer. Blood glucose monitoring was discontinued in 12 of the 42 patients 3e16 days prior to their death. In a further six, monitoring was reduced or discontinued within 24 hours prior to death, but rationales for treatment decisions were not provided. Additionally, it was not stated whether the guidelines McCoubrie et al.8 developed as a result of the audit had been pilot-tested or evaluated. The guidelines include recommendations for referral to specialist palliative care services and/or a diabetes team, relaxing dietary control, and aiming for a blood glucose level between 10 and 20 mmol/L. The other papers revealed differences in the recommended blood glucose targets and ranges, although the most consistent recommendation was to aim for a range between 6 and 15 mmol/L and freedom from symptoms.5,6 However, if hyperglycemic symptoms occurred, a range between 10 and 20 mmol/ L was recommended as a symptom management strategy.5e8 Poulson5 identified some of the specific challenges of managing diabetes and advanced disease. She suggested that Type 1 diabetes requires closer monitoring than Type 2 and asserted that the degree of control depends on the stage of the cancer; she indicated that less tight control is required when patients have a limited prognosis. Poulson recommended that active patients with advanced cancer enjoying a satisfactory quality of life must be monitored closely enough to ensure that their quality of life and physical activity are not compromised. However, she did not indicate how each of these realms was assessed. Relief from hyperglycemic symptoms (usually present when blood glucose is above 15 mmol/L), such as polyuria, thirst, and lethargy, can be achieved with insulin.6e9 However, Poulson5 questioned whether continuing


Diabetes Management in Palliative Care Patients

insulin might prolong life unnecessarily rather than enhance the quality of life, and suggested that little is known about using insulin for patients with advanced disease. The remaining 14 papers provided information about the pathophysiological changes in diabetes due to advanced terminal disease. A conference paper that described a retrospective audit of the medical records of 18 palliative care patients (diagnoses not provided) with Type 2 diabetes suggested that blood glucose levels do not rise, even when usual diabetic therapy is stopped.7 The authors did not state the average blood glucose level prior to ceasing diabetic therapy. Their recommendations, to manage diabetes-related symptoms while limiting unnecessary interventions, were consistent with the aims of palliative care: that is, relieving suffering.10 Several key palliative care textbooks were reviewed.11e13 However, diabetes was only mentioned briefly and no diabetes management guidelines for people with advanced disease were provided. Limited reference has been made to the incidence of diabetes in patients with advanced disease. It has been suggested that the incidence is higher than in the general population due to a number of factors, including the use of corticosteroids and other medications likely to cause altered blood glucose levels (such as diuretics and octreotide), the age range of palliative care patients, and the effects of malignancy.6 The existing literature is characterized by a paucity of evidence-based clinical management guidelines and minimal insight into the incidence of diabetes in patients receiving palliative care. Doctors’ and nurses’ experiences of caring for patients with diabetes and advanced disease have not been comprehensively enunciated. Thus, existing recommendations for managing diabetes in patients with advanced disease appear to be guided by individual expert opinion, rather than evidence. With this background, a study was designed to describe the current practices of doctors and nurses caring for patients with diabetes and advanced disease. The project was guided by the following questions: 1. What are the clinical experiences of doctors and nurses caring for patients with diabetes and advanced disease?

277

2. What decision-making criteria do doctors and nurses use to assess and manage glycemic control in patients with advanced disease? 3. Does the management of people with diabetes and advanced disease vary according to diabetes type? A two-phase study was conducted and consisted of focus groups and a cross-sectional survey. The Human Research Ethics Committee of St. Vincent’s Health Melbourne, Australia approved the research.

Methods Focus Groups Given the limited evidence for managing diabetes in palliative care patients, focus groups were undertaken with doctors and nurses to explore their experiences of caring for people with both diabetes and advanced disease. Focus groups are a recognized research methodology to explore attitudes and beliefs through discussion in a facilitated group setting.14,15 Two focus groups were conducted, one with metropolitan participants from Melbourne, Australia and one with participants from regional areas of Victoria, Australia to enable a broad range of experiences to be explored. Sample. A purposive sampling was chosen, with the aim of inviting experienced participants to be able to discuss the topic of diabetes and palliative care.15 The participants were endocrinologists, palliative care doctors, nurses with at least 3 years experience in palliative care, general practitioners (GPs), and diabetes nurse educators (DEs) from Victoria, Australia. Endocrinologists and DEs were identified by one of the investigators, a DE, who had access to an accurate listing for these practitioners. The other investigators were experienced in palliative care and identified services from which palliative care doctors and nurses could be recruited. Five regional and five metropolitan services were included to ensure representation from inpatient and community (home-based) settings, and to accommodate metropolitan and regional experiences. The managers of these services were contacted via telephone to inform them of the project and

278

Quinn et al.

seek their assistance with recruitment by promoting the project to their staff and facilitating attendance at the focus groups. A plain language statement explaining the project was mailed to those managers who agreed to assist, as well as to the potential participants. The managers canvassed interest among their staff in participating in the focus groups by raising it in team meetings. Doctors and nurses who were interested in attending the focus groups made initial phone contact with one of the researchers. Data Collection. Participants completed consent forms before any discussion commenced and were reminded that the discussion was being audiotaped for later verbatim transcription. The questions (Appendix 1) were developed by the researchers based on the literature review findings and their joint experiences in diabetes and palliative care. A moderator guided the discussion using a series of questions, and two assistants recorded participants’ comments but did not take part in the discussion. The assistants were particularly asked to note nonverbal behaviors, such as nods of agreement and body language, that might suggest interest or disagreement, which could not be captured on tape.15 Data Analysis. Debriefing sessions were held with the moderator and assistants immediately after each focus group. Focus group discussions were transcribed from the tapes. Data were analyzed using the framework method,16,17 which involves a five-stage process of becoming familiar with the data by reading the transcripts and field notes, and listening to the tapes to recognize recurring words, topics, and phrases. This process enabled a content framework to be identified and indexed. Emerging themes were charted and mapped in order to interpret the findings. Relevant quotations that expressed the essence of each major theme were selected from the transcriptions and were coded for future reference. The major themes identified from the transcribed focus group data were compared with the scribes’ observations and field notes, lending credibility to the findings.18,19 Two researchers independently undertook the analysis to limit researcher bias and then met to discuss and confirm their findings.


Cross-Sectional Survey A questionnaire was selected as an appropriate method to extend, complement, and confirm the results obtained from the focus groups and allow an opportunity to gain greater insight into the management strategies that doctors and nurses utilized when caring for patients with diabetes and advanced disease. The research team developed the questionnaire specifically for the study, based on the findings from the literature review and focus groups. A panel of experts from palliative care and diabetes were asked to review the questionnaire to establish face and content validity, but the questionnaire was not formally validated. The questionnaire consisted of seven main domains (Appendix 2) and had 27 Likert scales, three multiple-choice questions, and 14 open questions that were designed to give the participants the opportunity to make comments and/or provide details about issues they considered relevant or that were not covered in other questions. A copy of the questionnaire is available on request from the authors. Sample. Ten palliative care services (homebased and inpatient) were approached to canvas their interest in participating in the project. Some of these services also participated in the focus groups. Five regional and five metropolitan palliative care services agreed to participate, and provided total numbers of nurses and doctors within their services so that the appropriate number of questionnaires could be sent to them. In addition, each of the service managers was asked to provide names of GPs within their region who could be approached for participation. The questionnaires were mailed to the endocrinologists and DEs as identified by one of the researchers and distributed by the palliative care service managers to the palliative care participants. Data Collection. A total of 167 questionnaires were distributed. The questionnaires were selfadministered and anonymous. A covering letter explaining the research was attached to each questionnaire. Completion and return of the questionnaire was taken as consent to participate. Questionnaires were coded to identify whether they were from metropolitan or regional services, whether they were from



inpatient or community settings, and which health professional group the respondent belonged to. Individual doctors and nurses could not be identified from the code. Reminder letters and questionnaires were mailed to all potential participants 2 weeks after the first questionnaires were distributed because there was no way of knowing who had already responded. Both mail outs included stamped return addressed envelopes to encourage return of the surveys. Data Analysis. Data were entered into a spreadsheet, and descriptive analysis was undertaken. Codes identifying the location of the service, community, inpatient, metropolitan, regional, nurse, or doctor were also entered to calculate response rates from each location. All entries were checked by one of the researchers and a research assistant. Percentages were calculated for each response option for the Likert questions.

Results Focus Groups Participant Demographics. Table 1 provides a summary of participants’ demographic details. The metropolitan group consisted of nine participants, and the regional group consisted of eight participants. Participants included two endocrinologists, four regional nurses, five metropolitan nurses, two regional doctors, two DEs, and two metropolitan doctors. The gender difference by discipline reflects that of the state of Victoria, Australia. Surprisingly,

279

no palliative care doctors or nurses had undertaken any ongoing diabetes education. Lack of Guidelines for Practice. The participants agreed there was a paucity of established guidelines or evidence-based protocols to guide management of glycemic control in patients with advanced disease. They were concerned that the lack of such guidelines resulted in a ‘‘trial and error’’ approach to care that could result in less than optimal management. Stage of Illness and Blood Glucose Management. The participants unanimously agreed that diabetes management practices changed according to the disease stage. They indicated that blood glucose levels needed to be more tightly controlled and monitored in patients with a prognosis of some weeks or months, who were not anticipated to die within days, particularly patients who experienced symptoms such as dehydration and thirst or who became ketotic. Several participants commented that controlling hyperglycemia during the last weeks or months of life can reduce the severity of the associated symptoms. One doctor said, ‘‘A once-a-day injection [insulin] may be quite advantageous as it controls symptoms and improves quality of life in those remaining months or weeks, in that patient.’’ Establishing blood glucose targets for patients during the last weeks of life seemed unclear and based on previous experience. As one nurse said, ‘‘I think a lot of people do feel better if their glucoses are down around 10 or a little bit below rather than up around

Table 1 Focus Group Participant Demographics Discipline Age Range (y) Medicine

40e49

Nursing

27e55

Gender

Qualifications Post Initial Registration

Role

Male ¼ 5, female ¼ 0

Grad Dip Palliative Endocrinologist ¼ 1, palliative Care ¼ 3, MD ¼ 1, Master care physician ¼ 2, regional GP ¼ 2 Medicine ¼ 1 NUM ¼ 1, ANUM ¼ 2, CNC ¼ 5, Female ¼ 11, male ¼ 0 MN (palliative care) ¼ 1, Grad Diabetes CNE ¼ 2, Team Dip Palliative Care ¼ 2, Grad leader ¼ 1 Cert Diabetes Education ¼ 2, Grad Cert (Palliative Care) ¼ 2, Grad a Dip Education ¼ 2; 1

One participant declined to provide demographic data. ANUM ¼ Associate Nurse Unit Manager; Grad Dip ¼ Graduate Diploma; CNC ¼ Clinical Nurse Consultant; CNE ¼ Clinical Nurse Educator; MN ¼ Master of Nursing; Grad Cert ¼ Graduate Certificate; NUM ¼ Nurse Unit Manager. a Missing data.

280

Quinn et al.

15. That is my experience with patients, that’s what they tell me.’’ Several participants expressed difficulty determining whether symptoms during the last weeks of life, such as breathlessness, pain, drowsiness, dehydration, and confusion, were caused or exacerbated by abnormal blood glucose levels and contributed to increased distress and discomfort or were due to the disease process. One metropolitan nurse described the practice in her setting where blood glucose was routinely monitored four times a day for all patients with diabetes newly admitted to the service. However, there was some ambiguity in the plan of care following the initial monitoring. Sometimes, patients’ blood glucose levels were regularly recorded but these results were not acted on unless they were over 15 mmol/L. Another nurse said they plan care based on the symptoms experienced by the patient rather than setting rigid blood glucose targets. However, participants indicated that the level of blood glucose control could be considerably relaxed and testing ceased when patients are imminently dying (within days). A community regional nurse said, ‘‘The management of diabetes fades into the background when someone is imminently dying.’’ Most participants said that their management of patients expected to die within days was guided by the symptoms experienced by the patient rather than definitive, prescriptive blood glucose targets that required frequent blood glucose recordings. Participants discussed the challenge of balancing symptom control and patient comfort with a desire to limit invasive investigations, which is the goal of palliative care. Blood glucose monitoring was discussed at length, particularly the degree of discomfort caused by fingerpricking and the value of monitoring blood glucose in a patient imminently dying within days, without the benefit of any evidence to support the practice. A doctor expressed his difficulty estimating the benefit of actively treating symptoms versus striving to maintain the patient’s comfort: ‘‘. there is that tension between uncomfortable intervention [fingerpricking] but avoiding discomfort from something that can be improved [diabetes control].’’ Several participants felt blood glucose monitoring could be invasive and/or painful in


patients expected to die imminently from advanced disease. However, one nurse put forward a contrary contention when she stated, ‘‘We are pretty invasive in other ways with injections, syringe drivers, bowel care. How invasive is glucose monitoring?’’ A GP suggested that urine testing could be considered rather than frequent fingerpricking to determine blood glucose levels in patients who were imminently dying within days. He stated that while urine testing is less accurate, it might detect blood glucose levels greater than 10 mmol/L, which could then be followed up with a blood glucose test. Issues Specific to Type 1, Type 2, and Steroid-Induced Diabetes. All participants agreed that the different types of diabetes raised different management issues. They suggested that patient and family attitudes toward dietary control might vary. For instance, patients with long-standing Type 1 or Type 2 diabetes would ideally have been on an eating plan and may express concern over a more relaxed approach as their condition deteriorated. One doctor commented that some doctors and nurses consider people with Type 1 diabetes to have different issues than people with Type 2 diabetes, as outlined in the following quote: ‘‘We go through all these complex negotiations for people who are on insulin, who have to be monitored and treated, and yet for people who are on oral hypoglycemics, it is light years different and yet the issues are pretty much the same.’’ Participants appeared to be less confident managing patients who require insulin and determining when or how insulin is required for patients with advanced disease. Participants suggested that, in the general medical community, diabetes is not considered to be a lifethreatening condition and Type 2 diabetes is considered easier to manage and has less serious consequences than Type 1 diabetes. A palliative care nurse who worked in a community palliative care service described one example of the management strategy changing according to the type of diabetes. She said they continue to test blood glucose when patients have steroid-induced diabetes in order to recognize hyperglycemia, respond appropriately, and minimize the risk of hyperglycemic complications: ‘‘It would be a very simple sliding scale



281

previous 4 years. The palliative care doctors and nurses acknowledged they infrequently referred patients to an endocrinologist or DE and agreed there was opportunity for increased collaboration between the specialties. Regional clinicians reported they had limited access to education, support, and services. Participants spoke of the challenges of geographical isolation for patients once they were commenced on insulin. As one regional nurse said, ‘‘I think from a practical point of view . you have a patient at home and start them on insulin, you haven’t got the luxury of having a diabetes nurse educator come. particularly if they live an hour away on a farm.’’ Web-based resources were recommended as being especially useful for regional areas and clinicians working in isolation.

of three times a day Actrapid if their blood glucose is ever over 15, or something like that.’’ Participants discussed the effects of corticosteroids, such as masking signs of infection due to the anti-inflammatory effect of the medication, and exacerbating hyperglycemia in people with diabetes. Participants agreed that regular blood glucose monitoring was justified when corticosteroids were used, but they did not agree about the frequency or the most effective monitoring method. Communicating With and Educating Patients and Families. Participants acknowledged carer burden for families supporting people with diabetes and advanced disease. The following comment by one regional nurse identifies the distress that carers often deal with as part of their support role: ‘‘.a relative who has got a fairly high anxiety level and they are just hanging in there looking after that person, you add that sort of stuff [insulin] into it and they will just tip over the edge.’’ Participants discussed the challenge of supporting the patient and his or her family when they participate in making decisions about diabetes management changes, particularly when diabetes has been a major part of the individual’s life, often for a long period of time. One community nurse highlighted this issue: ‘‘.it is important that they don’t see cessation of insulin as kind of being an abandonment of all care.’’

Cross-Sectional Survey Respondent Demographics. Sixty-seven of the 167 questionnaires were returned, which represents a response rate of 40%. Table 2 shows responses per professional discipline. Responses were received from palliative care doctors and GPs in two of the five metropolitan services and from GPs in four of the five regional areas. A minimum of two nurses responded from each of the 10 participating services. It was unclear whether the low response rate from doctors in metropolitan areas reflected a lack of contact with patients who have diabetes and advanced disease, a lack of interest in this area of practice, unwillingness to participate, or whether it was attributable to other reasons.

Lack of Referral to Expert Professional Colleagues. The endocrinologists were rarely involved in caring for patients with advanced disease. One endocrinologist had received only one referral from palliative care services in the

Table 2 Response Rate per Health Professional Group (HPG) and Demographic Detailsa

Health Professional Group Endocrinologist Palliative care physician in a hospital GP metropolitan GP regional Registered Nurse metropolitan Registered Nurse regional a

Response Rate % of n ¼ 67

Mean Age in Years

Gender per HPG

Mean Year First Registered to Practice

Mean Years Experience in Palliative Care

Postgraduate Qualifications

10 7

45 46

M: 4, F: 3 M: 3, F: 2

1980 1981

0 9.4

6 4

4 15 39 24

54 50 43 45

M: M: M: M:

1974 1981 1985 1983

10.3 7.6 6.4 7.0

1 7 17 10

Three respondents did not provide demographic details.

3, 8, 2, 0,

F: F: F: F:

0 2 24 16

282

Quinn et al.

Table 3 provides a summary of the key findings from the questionnaire. For example, one-third of participants (33%) reported that they often lacked confidence to care for patients with diabetes and advanced disease, but they did not refer patients to diabetes experts. Endocrinologists and DEs stated they never received referrals to see patients with diabetes from palliative care experts, which confirmed the reported lack of referrals by palliative care doctors and nurses. Despite evidence that corticosteroid medicines increase the risk of diabetes,6 more than one-third (36%) of the respondents did not know how often patients on corticosteroid medicines had abnormal blood glucose levels, although approximately half (51%) stated they encountered steroid-induced diabetes in 25%e50% of patients. The greatest management challenge was ensuring patient comfort and limiting the number of intrusive investigations. One regional doctor wrote, ‘‘.deciding on when to be less active about blood glucose targets and the appropriate level of intervention is the real issue.’’ Over half (54%) of the respondents reported that they initiated an individualized approach to managing diabetes during the


last weeks of life. The approach was based on the patient’s blood glucose levels, nutritional status, and conscious state, the wishes of the patient and his or her family, and the overall medical history, and the patient was treated to control symptoms rather than achieve rigid blood glucose targets. Ninety percent of participants considered blood glucose monitoring to be important in end-of-life care and a variety of blood glucose ranges were suggested as appropriate in the last weeks of life (Table 4). However, only 6% of respondents initially tested blood glucose regularly for a limited period, between 24 and 48 hours from the time of admission to the service, and then adjusted the testing frequency depending on the results. The majority (86%) of participants aimed for a blood glucose level between 5 and 20 mmol/L for patients with some weeks to live. Managing Diabetes During the Last Days of Life. The approach to blood glucose monitoring differed for patients who were in the last days of life from that for patients with a longer prognosis. More than one-third of participants (37%) stated that they ceased monitoring blood glucose levels during the last days of life; however,

Table 3 Summary of Key Findings from Questionnaire Responses Practice Domain Experience of doctors and nurses with diabetes in end-of-life care

Confidence of doctors and nurses discussing end-of-life issues with patients with diabetes and their families Referrals received by endocrinologists and DEs from palliative care Perceived competence to diagnose diabetes at end of life Doctors and nurses self-rating on managing diabetes at end of life Awareness in doctors and nurses of how often patients on corticosteroid medication had abnormal blood glucose levels Doctors and nurses’ perceptions of percent of patients with corticosteroid-induced diabetes in end-of-life care Attitude of doctors and nurses to Type 1 and Type 2 diabetes

Formal diabetes education undertaken by the doctors and nurses after initial qualifications Regularity of review of diabetes for palliative patients Doctors and nurses’ satisfaction with diabetes management

Key Results 34% regularly cared for patients with diabetes and advanced disease 55% sometimes cared for patients with diabetes and advanced disease 11% were uncertain how often their service provided care to patients with diabetes and advanced disease 30% confident 33% lacked confidence None 48% confident 40% not confident 82% as average to good 36% unaware 25%e50% 49% considered Type 1 more life threatening than Type 2 22% considered them equally life threatening 29% unsure 13% yes 87% no 25% reviewed daily 42% reviewed weekly 42% rated satisfaction less than 5 on a scale of 0e10



Table 4 Recommended Blood Glucose Range During the Last Weeks and Days of Life

Blood Glucose Level Range (mmol/L)