Quality of life in terminal cancer patients and their relationship with different models of home palliative care.

Agra Y, Sacristán A, Pelayo M, Fernández J.
-Yolanda Agra: Family Practice, PhD. Primary Health Center. Zone 1. Imsalud Madrid.
Plaza del Doctor Laguna, 6-4º B. 28009. Madrid. Spain
e-mail: yoagra@santandersupernet.com
Fax= 34-915045401
-Antonio Sacristán: Oncologist, PhD. Palliative Care Unit. Zone 4. Imsalud. Madrid
Calle Santa Isabel, 15-2º C. 28012. Madrid. Spain
e-mail: conchapy@hotmail.com
-Marta Pelayo: Family Practice. Primary Health Center. Zone 7. SERVASA. Valencia
Urbanización Sierramar 104. Picasent. Valencia. Spain
e-mail: mpalva@terra.es
-Julia Fernandez: Family Practice. Palliative Care Unit. Zone 2. Imsalud. Madrid.
Calle Luis Velez de Guevara, 7-1º izda. 28012 Madrid. Spain
e-mail: julia.fernandez@guanodoo.es

Manuscript correspondence:
Dr. Yolanda Agra
Plaza del Doctor Laguna, 6-4 B. 28009. Madrid. Spain
e-mail: yoagra@tiscali.es

Abstract

Objectives: to evaluate if the terminal cancer patients treated by a Palliative Care Unit have better quality of life than the patients treated with standard care. To determinate differences in control symptoms, drug prescription and place of death.
Methods: prospective quasi-experiment study undertaken in Zone 4 Health District in Madrid. Terminal cancer patients referred to the Hospital to be treated by the Palliative Care Unit (PCU) or by the General Practitioners in the Primary Health Centers (PHC). The main outcome was quality of life assessed by the Rotterdam Symptom Check List (RSCL) and the Hospital Anxiety and Depression Scale (HAD). Statistical differences were evaluated by MANOVA and chi-square. Drug prescription and place of death were also assessed.
Results: a total 165 patients were treated by PCU and 54 by PHC . The patients treated by PCU presented better scores in global (IC 99%: 17.94,1.02) and pain (IC 99%: 18.85, 0.12) scales of the RSCL. The percentage of patients with psychological symptoms decreased in PCU across the study (p < 0.003). Drug prescriptions and place of death were similar in both groups.
Conclusions: the patients treated by the PCU had a better general perception of health than patients treated by PHC. The fact than drug prescription and place of death were similar between groups and the high number of dropouts across the study, it make sense to take the results cautiously.
Key words:
Quality of life, palliative care, terminal care, terminally ill, neoplasms.

Introduction

The results in sanitary oncology intervention are assessed, in general, in terms of mortality and survival. The main objectives of the Health System are to prolong survival and allow a satisfactory quality of life. This aim represents an important economic cost both in research and development of new and expensive methods of diagnosis and in cancer treatment.
Despite all the new technological advances in the last few decades the epidemiological data shows that the incidence of cancer and cancer mortality are increasing. Now 50% of all patients with cancer die as a result of the progression of their illness1
Cancer is the second cause of mortality in Spain, with 89.000 deaths per year2.
In most of these patients it is possible to identify a “terminal phase”, defined by the progression of the illness, presence of symptoms, a bad prognosis in a short time and the lack of specific oncologycal treatment. Terminal cancer patient care has to be orientated to the control of symptoms and to provide a reasonable quality of life1. The present tendency in health policy in Europe3 is to organize palliative care at home and to transfer this responsibility to the primary care setting whenever there exists the basic conditions to do so (if the symptom control is the same as in the Hospital, the home is adequate and with the presence of a care giver). This is expected to produce a better quality of life for the patients and lower the cost to the Heath System. However the effectiveness of the palliative care programmes at home, for these patients, are not clearly stated4,5.
Now, in the Spanish Health System, palliative care can be offered both in hospitals and at the patient´s home. In the second case, the care can be offered by general practitioners (or family physicians) working in a primary health centre (PHC), specialized professionals working in a palliative care unit PCU or both working in a coordinated follow-up programme6. This situation, similar throughout the European Union7, has clear advantages but some problems. The possibility both of adequated coordination between doctors and also the possibility that the patients do not all receive the same quality of care. Although there is not yet conclusive data about the most effective care for terminal cancer patients, some studies have shown benefits in symptom control by specialist palliative care teams7,8.
In theory the responsability of the palliative care has to be assumed by the professionals with the best results in terms of symptom control and quality of life.
In Zone 4 of IMSALUD (Health Council) in Madrid, Spain, there has been since 1991 a Palliative Home Care Programme (PHCP) for terminal cancer patients (TCP)9 that has a palliative care unit (PCU) made up of two doctors and two nurses specialized in palliative home care. The main objectives of this team are: to take care of terminal cancer patients and to support general practitioners in the symptom control of these patients.
General practitioners and professionals of the PCU follow the PHCP recommendations. So, the patients can be cared for by family physicians working in a PHC, by the PCU or by both at the same time.
According to a previous pilot study, patients controlled by PCU, in our Zone, presented better symptom control than patients controlled by their family doctors. For this reason the present study was undertaken to evaluate if the terminal cancer patients controlled by a PCU had a better quality of life than patients controlled by family physicians with conventional care.
A secondary objective was to assess differences in symptom control, drug prescription and the place of death.

Materials and Methods:

This prospective quasi-experiment study was carried out at Zone 4 of the Imsalud (Health Council) in Madrid that has 537.000 inhabitants. The PHCP9 of this Zone for terminal cancer patients at home describes the inclusion criteria, home visit planning and general guidelines for the care and treatment of these patients.
The sequence of entry and the follow-up of patients in the PHCP is the following: in most cases the reference Hospital informs the PCU about the patients that fulfil the inclusion criteria of entry in the programme. This information is sent to the family physician responsible for the patient, who will then decide whether to take the care of the patient or to delegate this care to the PCU. In the second case, the patient is visited by a physician and a nurse from PCU in the first 48 hours to prescribe the treatment needed and to plan the follow-up visits. If the patient is in the care of the PHC the family physician in charge of the patient, will decide the timing of the home visits .
To assess the objective of this study the patients were selected, from April 1, 1999 to March 1, 2000, consecutively if they had fulfilled the following criteria: terminal illness (with clearly documented terminal disease and a life expectancy less than 6 months according to the hospital release report), age > 18; knowledge by the family of the terminal illness and having a caregiver (a person living with the patient providing his/her basic care at home) and an informed consent to take part in the study. Patients with Karnofsky Performance Status10 (KPS) < 20 or Pfeiffer11 test with more than three mistakes or patients controlled by PHC and PCU together were excluded.
On hundred and sixty two family physicians were invited to participate in the study, without knowledge of its objectives, through specific meetings and personal letters. They should give information about the new patients in the programme and facilitate access to their clinical data.
The quality of life, the main variable of this study, was assessed by the Spanish version of the Rotterdam Symptom Check List (RSCL) whose validity, reliability and responsiveness to clinical changes was shown in previous studies12,13.
This questionnaire comprises 39 items which are grouped into four scales: physical symptoms (23 items), psychological symptoms (7), activities in daily life (8) and overall quality of life (1). The answers are all of the Likert type in four categories, except for the overall quality of life scale with seven categories. The scale of physical symptoms is made up of four subscales: pain, fatigue, gastrointestinal and chemotherapy symptoms. The scales and subscales are transformed into a rating ranging from 0 to 100. The highest rating obtained is linked to the worst health prognosis and the lowest rating with the best for all the scales and subscales. On the activity scale, on the contrary, the highest rating was linked to the best functional state14.
The psychological symptoms were assessed by the Spanish version of the Anxiety and Depression Scale (HAD)15. This questionnaire consists of 14 items, which are first standardized from 0 (the most favourable state) to 3 (the least favourable). Seven of these items are then added together to produce the anxiety scale, and the other seven constitute the depression scale. Thus each scale ranges from 0 to 21. A score of 11 or more is regarded as apotential clinical case, a score between 8 and 10 is regarded as boderline, and one of ≤ 7 regarded as normal16.
To know the pharmacological treatment the drugs assessed were: analgesic drugs (non opioid), opioid, corticosteroids, benzodiazepines, antidepressants, anticonvulsants, neuroleptics, laxatives and antiemetics
Other variables measured were: age, sex, KPS, location of the primary tumour and annual family income (I: < 6,000 €; II: ≥ 6,000 and < 15,000 € ; III: ≥ 15,000 and < 24,000; IV: ≥ 24,000 € ).
The questionnaires, RSCL and HAD, were administered at the patient´s home in the first 48 hours of admission to the Programme, 7 and 15 days later by a previously trained physician interviewer. For items with no data available, the author´s recommendations from the original version were followed14: the average rating of the scale was included as long as the patient had answered at least 50% of the items on the scale.
The resting variables, including drug prescription and place of death (home, hospital, etc), were registered from standarized clinical records.
The causes of dropouts considered were: relocation to another district, hospitalization, worsening of the clinical situation and death.
To facilitate analysis of the data, the KPS was grouped in three categories (30-40, 50-60 and 70-80) and the social class in two (I-II and III-IV). The data were analysed with the SPSS statistical package in Windows, version 6.1.117.
Sample size was calculated according to previous studies with the RSCL13. At least 56 patients, in each group, were considered necessary to detect a difference of 10 points in the scales between groups with a standard deviation= 20.6, error a=0.05 and b =0.20 for a unilateral tets18.
To assess differences, in scale ratings in the second and third visits, between groups the MANOVA was used. The significance of age, in the model, was evaluated by multivariate linear regression16,18.
The assess differences , in each group, effect size (ES) was calculated between the first and the other visits. This is a standardized measure to allow the detecting of changes in health status independently of the statistical significance. The effect size ≥ 80 is considered very high, around 50 moderate and <20 very small19,20.
Differences in the proportion of patients with mood distress (anxiety and depression) according to HAD, were evaluated by Chi-square test between groups in the second and third visit and in each group between the first and the other visits .
Bonferroni correction was performed with a =0.01.
The drug prescription and place of death were explored by simple descriptive statistics.

Results

A total of 215 patients were included in the study, of which 161 were treated by PCU and the rest by the selected PHC. The clinical and personal characteristics of the patients are shown in table 1. The average age of the patients was 70.6 years (standard deviation: 10.4; range: 37 to 95), the majority were men, with lung cancer as the most frequent tumour. A total of 88.8% of the patients had a KPS ≤ 60. A total of 65% of the subjects stated that their total annual family income was lower than 15,000 euros. There were no statistically significant differences between the PCU and PHC patients for these variables nor for the rates of metastases that were 76.7% and 76.9%, respectively. The location of the metastases was similar in both groups, with liver as prime location (27%, 22.6%, respectively ) followed by bone (25%, 21.2%). The average survival in days was slightly higher in PHC patients (m: 74.3; sd: 73.4) than in PCU patients (m: 53.5; sd: 53.6).
Of the total patients who started the study, 169 patients (79%) met the necessary requirements to complete the second questionnaire and of these, 94 (44%) were able to complete the third one. The main reason for the dropout rate was due to death (66.7%), followed by hospitalisation or change of address (22.3%,) and a deterioration in the functional state of the rest. The percentage of the dropout rates and their causes were similar in PCU and PHC (57% and 54% respectively). The dropouts during the course of the study did not produce significant differences between the two groups in terms of age, sex, functional state or type of tumour. The characteristics of patients unable to complete the second questionnaire were similar in both groups though with higher scores (worse health perception) than patients in study (table 2). As can be seen in table 3, RSCL scores were similar in both groups at the beginning, except for the quimiotherapy scale. In the second visit, PCU patients had better health perception (lower scores) in the global (IC 99%: 17.94; 1.02, p=0.01) and pain (IC 99%: 18.85; 0.12 p=0.01) scales. These results didn’t show any change when age was included in the model. In the third visit there was a tendency to better scores in PCU patients but with non significant differences (table 3).
When each group was analysed separately, the effect size was important or moderate (effect size around 50) for the physical scale between the first and second visit (ES= 0.59) and for the pain (ES= 0.55) and gastrointestinal scales (ES= 0.50) between the first and third visit in PCU. Effect size for the rest of scales in PCU and for all those measured in PHC was small (ES less than 0.45).
Total percentage of patients with anxiety or depression measured by the HAD questionnaire was 11% and 32% respectively. There were no relevant differences between groups (table 4) but the percentage of patients with anxiety or depression decreased significantly between the first and follow-up visits in PCU (p ≤ 0.003) and not in PHC (p= 0.54).
In table 5 description of pharmacological profile refers to those prescribed between the first and the second visit. Narcotics were used in 80.5% of which 42.8% was morphine. The maximum dose of morphine was 460 mg in PCU and 320 mg in PHC. PCU prescribed more acetaminophen, benzodiazepines and steroids; PHC prescribed more narcotics and anti-inflammatories.
Death at home was 58.8% in all patients, 56.6% in PCU and 65.8% in PHC.

Discussion

Data of this study shows a better health perception of PCU patients in global and pain scales of RSCL and a decrease in psychological symptom measured in the HAD questionnaire in comparison to PHC patients. Given that the clinical characteristics of patients (age, sex, kind of tumour or KPS) didn’t change and were not different between groups, this cannot adequately explain the differences found. Our results are in accordance with other publications which show an improvement in the quality of life in patients controlled under specific units of palliative care21. However, when comparing some care model results of the majority of the studies, there is not a relevant difference between specialized or standardized care22,23, or if there is one, it is in favour of better pain control in specialized teams5,24.
As differences were detected in the first week and were not maintained throughout the study, an insufficient number of subjects might be the cause in part, due to a low number of PHC patients refered from the general practitioners and dropouts that reduce the statistical power to detect differences. It is known that measuring results in oncological palliative care is difficult due to the dropout of patients25. In our study, 44% of subjects are still in the programme 15 days after beginning, with 90% of the dropouts due to death or deleterious functional state. This situation is also described in other studies21,26, and has been the reason to choose a short period of follow-up even though this could facilitate a memory bias and make score responses towards a better health perception.
Low participation of doctors in PHC (20%) might be in relation to a professional feeling of being inspected in their clinical work or to a low degree of motivation.
The questionnaires to measure quality of life in prospective studies need to have adequate psychometric properties to detected changes in state of health when performing different intenventions18. Spanish version of RSCL has shown validity, reliability and sensibility to changes in previous studies with terminal cancer patients of similar characteristics to those included here12,13. RSCL has the property to measure symptoms, a fact of much utility in the treatment of cancer patients, it is short and easy to administer. HAD has been useful in palliative care since it is accepted by patients and permits detection of psychological impairments susceptible to treatment27.
Symptoms referered by patients are similar to other series both for anxiety and depression28,29.
Drug prescription is similar in both groups and in accordance to actual recommendations29 for the treatment of symptoms in these patients. In other studies a higher prescription of narcotics has been seen in specialized teams than in primary care3 a fact that has not been seen here. A reason for this difference could be the use of the same protocol in PCU and PHC. The adequacy of drug doses prescribed has not been analysed here and might be a determinant factor in symptom response.
The percentage of patients dying at home is higher than reported in other studies30. In Spain death at home seems to occur more frequently than in other European countries. The fact that women, principally responsible for care of these patients, have not yet reached the same labour incorporation level as in the rest of the European Union31, might justify in part this result. Nevertheless our rates are lower to those reported in home hospitalization32, where specialized personnel are available 24 hours a day to offer care.
Another possible cause of the dilution of differences might be the training in palliative care for the general practitioners, working in the Zone 4, in the previous years.
Primary care doctors need to be responsible for the palliative care of their patients in a multidisciplinary context 7,33, since some studies have shown the desire of these professionals to maintain this specific instruction 33,34. In this sense the objective would be not to determine an exclusive service of palliative care but to draw up plans to implement training programmes in palliative care for family practices to improve the quality of life in terminal cancer patients, maintaining the principle of fairness.
Recommendations to implement a sanitary measure of intervention need to be based on evidence of clinical trials8 but this has not been possible here due to planification measures taken in our place. Nevertheless aspects as "contamination" have not been very likely since doctors have not known the study objectives and patients at home have not known of the care offered to other patients.
This study shows an improvement in physical, global and pain scales of RSCL and a decrease in psychological symptoms during the study of terminal cancer patients treated by specialists in palliative care. However the fact other variables of interest in palliative care as place of death and drug prescription have been similar in both groups, and given the small number of patients that completed the study, it seems advisable to treat the results cautiously.
Acknowledgment
Study financed by the Ministry of Health (FIS expte: 99/0482)

References

1.Doyle D, Hanks G, Mac Donald N. Oxford Textbook of Palliative Medicine. 2nd ed. New York: Oxford University Press; 1997
2.Instituto Nacional de Estadistica. Movimiento natural de la población española. Madrid: INE; 1996.
3.Lamí U, Kosunen E, Kellokumpu-Lehtinen P. Palliative cancer care in two health centres and one hospice in Finland. Support Care Cancer 2000; 9:25-31.
4.Smeenk FWJM, van Haastregt JMC, de Witte LP, Crebolder HFJM. Effectiviness of home care programes for patients with incurable cancer on their quality of life and time spent in hospital: systematic review. BMJ 1998;316: 1939-44
5.Fordham S, Dowrick C. Is care of the dying improving? The contribution of specialist and non-specialist to palliative care. Family Practice 1999; 16:573-79)
6.Centeno C, Arnillas P, Hernansanz S, Flores LA, G’omez M, Lopez-Lara F. The reality of palliative care in Spain. Palliat Med 2000;14:387-94
7.McLaren G, Preston C, Grant B. Evidence based palliative care [letter]. BMJ 1999; 319: 1573
8.Higginson IJ. Evidence based palliative care. BMJ 1999;319:462-3.
9.Insalud Madrid Área 4. Programa de Cuidados Paliativos a Domicilio. Madrid: Dirección de Atención Atención Primaria Área 4. 2 ed: Insalud; 1997
10.Karnofsky DA, Abelmann WH, Craver LF et al. The use of nitrogen mustards in the pallliative treatment of carcinoma. Cancer 1948; 1:623-56
11.Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatrics Soc 1975; 23:433-41
12.Agra Y, Badía X. Spanish version of the Rotterdam Symptom Check List: cross-cultural adaptation and preliminary validity in a sample of terminal cancer patients. Psycho-Oncology 1998;7:229-39
13.Agra Y, Badía X. Evaluación de las propiedades psicométricas de la versión española del Rotterdam Symptom Check List para medir calidad de vida en pacientes oncológicos. Rev Esp Salud Pública. 1999; 73: 35-44.
14.de Haes JC, Olschewski M, Fayers P, Visser MRM, Cull A, Hopwood P et al. Measuring the quality of life of cancer patients with the Rotterdam Symptom Check List (RSCL): A manual. Northen Centre for Health Care Research (NCH). The Netherlands: University of Groningen; 1996.
15.Tejero A, Guimerá EM, Farré JM. Uso clínico del HAD (Hospital Anxiety and Depression Scale) en población psiquiátrica: Un estudio de su sensibilidad, fiabilidad y validez. Rev Psiquiatr Fac Med Barc 1986; 13:233-38
16.Staquet MJ, Hays RD, Fayers PM. Quality of life assessment in clinical trials. Methods and practice.Oxford New York: University Press.Oxford; 1998
17.SPSS [computer programme]. Version 6.1.2. para Windows en espanol: Microsft Corporation;1995.
18. Kleinbaum D, Kupper L, Muller K. Applied Regression Analysis and Other Multivariable Methods. 2nd ed. Boston: PWS-KENT; 1997
19.Hays RD, Anderson R, Revicki DA. Psychometric evaluation and interpretation of health-related quality of life data. In: shumaker SA, Berzor RA, editors. The International Assessment of Health Related Quality of Life: Theory, Translation Measurement and Analysis. Oxford: Rapid Communications of Oxford Ltd; 1995. P.103-11
20.Kazis L, Anderson J, Meenan R. Effect Sizes for interpreting changes in health status. Medical Care 1989; 27 Suppl: 178-89. 21.Cohen R, Boston P, Mount BM, Porterfield P. Changes in quality of life following admission to palliative care units. Pall Med 2001; 15:363-71
22.Eran J, Higginson IJ. Do specialist palliative care teams improve outcomes for cancer patients? A systematic literature review of the evidence. Pall Med 1998;12:317-32
23.Jordhoy MS, Fayers P, Loge JH, Ahlner-Elmqvist M, Kaasa S. Quality of life in palliative cancer care: results from a cluster randomized trial. J Clin Oncol 2001; 19: 3884-94
24.Salisbury C, Bosanquet N, Wilkinson EK, Franks PJ, Kite S, Lorentzon M et al. The impact of different models of specialist palliative care on patients´quality of life: a systematic review. Pall Med 1999; 13: 3-17.
25.McWhinney IR, Bass MJ, Donner A. Evaluation of a palliative cae service: problems and pitfalls. BMJ 1994;309:1340-42
26.Donnelly S, Rybicki L, Walsch D. Quality of life measurement in the palliative management of advanced cáncer. Supp Care Cancer 2001; 9:361-65
27.Holtom N, Barraclough J. Is the Hospital Anxiety and Depression Scale (HADS) useful in assessing depression in palliative care? Palliat Med 2000;14:219-20
28.Weiss S, Emanuel L, Fairclough D, Emanuel E. Unedrstanding the esperience of pain in terminally ill patients. The Lancet 2001; 357:1311-15.
29.WhitecarP, Jonas P, Clasen M. Managing Pain in the Dying Patient. Am Fam Phys 2000;61:755-64
30.Higginson I,Astin P, Dolan S. Where do cancer patients die? Ten-year trends in the place of death of cancer patients in England. Pall Med 1998; 12:353-63
31.Seasonally adjusted unemployment rate. http://europa.eu.imt/comm/eurostat/
32.Grande GE, Tood CJ, Barclay SI, Farquhar MC. Does hospital at home for palliative care facilitate death at home? Randomised controlled trial. BMJ 1999; 319:1472-5
33.Brow JB, Sangster M, Swift J. Factors influencing palliative care. Qualitative study of family physicians`practices. Can Fam Physician 1998; 44:1028-34
34.Shipman C, Addington-Hall J, Barclay S, Briggs J, Cox I, Daniels L, Millar D. Educational opportunities in palliative care: what do general practitioners want? Pall Med 2001; 15: 191-96

Table 1. Clinical and demographic characteristics of patients included
Variable PCU (n= 161) PHC (n= 54) TOTAL (n= 215)
Age: m, sd, (range)
Sex:
Men
Women
Kind of tumour:
Lung
Colon, rectum
Stomach
Other digestive
Breast
Liver
Prostate
Other
Unknown
KPS:
30-40
50-60
≥ 70
Social class:
< 15.000 €
≥ 15.000 €

70.7; 10.7 (37-95) 96 (59.6)
65 (40.4)

41 (25.5)
24 (14.9)
13 (8.1)
14 (8.7)
10 (6.2)
12 (7.5)
7 (4.3)
33 (20.5)
7 (4.3)

37 (23.0)
103 (64.0)
21 (13.0) 101 (62.7)
60 (37.3)
70.2; 9.5 (51-87) 30 (55.6)
24 (44.4)