- Email: [email protected]
Limitation of therapeutic effort in patients hospitalized in departments of internal medicine
Rev Clin Esp. 2018;218(1):1---6
Revista Clínica Española www.elsevier.es/rce
ORIGINAL ARTICLE
Limitation of therapeutic effort in patients hospitalized in departments of internal medicine夽,夽夽 R. García Caballero a,b,c,∗ , B. Herreros a,b,c , D. Real de Asúa a,c,d , S. Gámez e , G. Vega f , L. García Olmos g a
Grupo de Trabajo de Bioética de la Sociedad Espa˜ nola de Medicina Interna, Madrid, Spain Unidad de Medicina Interna, Hospital Universitario Fundación Alcorcón, Alcorcón, Madrid, Spain c Instituto de Ética Clínica Francisco Vallés, Universidad Europea, Madrid, Spain d Servicio de Medicina Interna, Hospital Universitario de La Princesa, Madrid, Spain e Hospital Universitario Infanta Cristina, Parla, Madrid, Spain f Hospital Universitario Quironsalud, Pozuelo de Alarcón, Madrid, Spain g Unidad docente multiprofesional de Atención Familiar y Comunitaria, Sureste, Madrid, Spain b
Received 26 June 2017; accepted 3 October 2017 Available online 2 December 2017
KEYWORDS Limitation of therapeutic effort; No cardiopulmonary resuscitation; Decision making; Internal medicine
Abstract Introduction: There is little information on the limitation of therapeutic effort (LTE) in patients admitted to hospital internal medicine units. Objectives: To describe the indicated LTE regimens in the departments of internal medicine and the characteristics of the patients who undergo them. Patients and methods: An observational, descriptive retrospective study was conducted on 4 hospitals of the Community of Madrid. The study collected demographic and comorbidity data and the LTE orders prescribed for all patients who died during a period of 6 months. Results: The study included 382 patients with a mean age of 85 ± 10 years; 204 were women (53.4%) and 222 (58.1%) came from their homes. Some 51.1% of the patients were terminal, 43.2% had moderate to severe dementia, and 95.5% presented at least moderate comorbidity. Some type of LTE was performed in 318 patients (83.7%); the most common orders were ‘‘No cardiopulmonary resuscitation’’ (292 patients, 76.4%; 95% CI 72.1---80.8), ‘‘Do not use aggressive measures’’ (113 patients, 16.4%; 95% CI 13.7---19.4) and ‘‘Do not transfer to an intensive care unit’’ (102 cases, 14.8%; 95% CI 12.3---17.7). Some type of LTE was performed in 318 patients
夽 Please cite this article as: García Caballero R, Herreros B, Real de Asúa D, Gámez S, Vega G, García Olmos L. Limitación del esfuerzo terapéutico en pacientes hospitalizados en servicios de medicina interna. Rev Clin Esp. 2018;218:1---6. 夽夽 The study is part of doctoral thesis project by Rebeca García Caballero. ∗ Corresponding author. E-mail address: [email protected] (R. García Caballero).
2254-8874/© 2017 Elsevier Espa˜ na, S.L.U. and Sociedad Espa˜ nola de Medicina Interna (SEMI). All rights reserved.
2
R. García Caballero et al. (83.7%); the most common orders were ‘‘No cardiopulmonary resuscitation’’ (292 patients, 76.4%; 95% CI 72.1---80.8), ‘‘Do not use aggressive measures’’ (113 patients, 16.4%; 95% CI 13.7---19.4) and ‘‘Do not transfer to an intensive care unit’’ (102 cases, 14.8%; 95% CI 12.3---17.7). Conclusions: LTE is common among patients who die in Internal Medicine. The most widely used regimens were ‘‘No CPR’’ and the unspecific statement ‘‘Do not use aggressive measures’’. The patients were elderly and had significant comorbidity, terminal illness and advanced dementia. © 2017 Elsevier Espa˜ na, S.L.U. and Sociedad Espa˜ nola de Medicina Interna (SEMI). All rights reserved.
PALABRAS CLAVE Limitación del esfuerzo terapéutico; Órdenes de no reanimación cardiopulmonar; Toma de decisiones; Medicina interna
Limitación del esfuerzo terapéutico en pacientes hospitalizados en servicios de medicina interna Resumen Introducción: Existe escasa información sobre la limitación del esfuerzo terapéutico (LET) en pacientes ingresados en unidades de hospitalización de medicina interna. Objetivos: Describir las pautas de LET indicadas en los servicios de medicina interna y las características de los pacientes que las reciben. Pacientes y métodos: Estudio observacional descriptivo y retrospectivo de 4 hospitales de la Comunidad de Madrid. Se recogieron datos demográficos, de comorbilidad y las órdenes de LET pautadas en todos los pacientes fallecidos en un periodo de 6 meses. Resultados: Se incluyeron 382 pacientes cuya edad media fue de 85 ± 10 a˜ nos; 204 eran mujeres (53,4%) y 222 (58,1%) procedían de su domicilio. El 51,1% eran enfermos terminales, el 43,2% tenían demencia moderada/grave y el 95,5% presentaban comorbilidad al menos moderada. En 318 pacientes (83,7%) se realizó algún tipo de LET, siendo las más frecuentes las órdenes de «no reanimación cardiopulmonar» (292 enfermos, 76,4%; IC 95%: 72,1-80,8), «no usar medidas agresivas» (113 pacientes, 16,4%; IC 95%: 13,7-19,4) y «no ingresar en unidad de cuidados intensivos» (102 casos, 14,8%; IC 95%: 12,3-17,7). Conclusiones: La LET es muy frecuente en los pacientes que fallecen en medicina interna. Las pautas más utilizadas son «no reanimación cardiopulmonar» y la expresión poco concreta de «no usar medidas agresivas». Los pacientes son de edad avanzada, con importante comorbilidad, enfermedad terminal y demencia avanzada. © 2017 Elsevier Espa˜ na, S.L.U. y Sociedad Espa˜ nola de Medicina Interna (SEMI). Todos los derechos reservados.
Background Numerous expressions have been used to refer to the limitation of therapeutic effort (LTE), but none of them accurately explain the concept. A widely accepted definition is ‘‘not applying measures disproportionate to the therapeutic goal (mainly healing, improving or relieving) to patients with poor vital prognoses and a poor quality of life’’.1,2 A disproportionate measure is understood as an imbalance between costs-charges and benefits for the patient. LTE attempts to prevent unnecessary suffering in patients in their final phase of life.1 To consider a patient a candidate for LTE measures, various scales have been implemented that include factors such as quality of life, comorbidity, risk and vital prognosis.3---6 LTE has been mainly evaluated in other health departments such as intensive care,7 palliative care,8 oncology,9 geriatric medicine10 and pediatrics.11 The departments of internal medicine treat a significant percentage of hospitalized patients, many of them in their final period of life.12 However, there is very little consistent information on the LTE measures to be applied to these patients,13,14 whose demographic and clinical characteristics
differ significantly from those of patients hospitalized in other departments. This study describes the LTE guidelines indicated in internal medicine departments and the demographic and clinical characteristics of the patients who receive these guidelines. This study is the continuation and extension of a previous project (Withholding and withdrawing treatment in patients admitted in an Internal Medicine ward), where we analyzed the LTE in a single hospital.15
Patients and methods This was an observational, descriptive, retrospective multicenter study on 4 hospitals of the Community of Madrid, which provide care to urban and rural populations. We included all those patients who died during their hospitalization; there were no exclusion criteria. The study lasted 6 months and was conducted according to good clinical practice criteria and the Declaration of Helsinki.16,17 We obtained the following variables from the medical records: age, sex, hospital stay, patient origin (institution/residence vs. home), cognitive impairment (according to the Clinical Dementia Rating scale),18 comorbidity
Limitation of therapeutic effort in hospitalized of internal medicine (age-adjusted Charlson index: absence of comorbidity [1---2 points], mild [3---4 points], moderate [5---6 points] and severe comorbidity [>7 points]19 ), terminal status criteria at admission (according to the Spanish Society of Palliative Care20 ), reason for hospitalization and cause of death. The analyzed LTE measures were no cardiopulmonary resuscitation (CPR) order, hospitalization in the intensive care unit, use of renal replacement therapy, blood-product transfusion, implementation of diagnostic tests, surgery, parenteral nutrition, chemotherapy or radiation therapy, mechanical ventilation or intubation and the nonimplementation of ‘‘aggressive measures’’. We recorded only those orders that explicitly appeared in the medical progress notes. We also recorded the time elapsed from admission to the LTE order, as well as the palliative sedation measures1 and the duration of sedation. The data were analyzed with SPSS, version 17.0 (IBM Corp., Armonk, NY). We calculated the 95% confidence intervals of the primary endpoint (LTE regimen). The qualitative variables are presented using frequency tables (percentages), and the quantitative variables are presented as mean and standard deviation. We analyzed the associations between the LTE orders, abstention from CPR and palliative sedation on one hand and the following variables on the other: degree of dementia, institutionalization, Charlson index, sex and age. To perform the analysis, we categorized the Charlson index using the median as the cutoff (8 points). The degree of dementia was grouped into 2 categories: absent/mild (grade 0---1 on the Clinical Dementia Rating scale) and moderate/severe dementia (grade 2---3). We calculated the odds ratio (OR), its confidence intervals and the significance level using the chi-squared test for the contrast of hypothesis.
Results The study analyzed 382 patients with a mean age of 85 ± 10 years; 204 were women (53.4%) (Table 1), and the majority (222, 58.1%) came from their homes. The mean hospital stay was 11 ± 13 days. Some 43.2% of the patients had moderate-severe dementia, 95.5% had at least moderate comorbidity, and 51.1% met terminal status criteria. There were 39 (10.3%) unexpected deaths, understood as those for whom there were no comments indicative of imminent death in the medical and nursing notes in the days prior to the death. Other general demographic and clinical characteristics of the sample are shown in Table 1. The causes of hospitalization and death are indicated in Table 2. An LTE measure was established for 318 patients (83.7%; 95% CI 79.3---87.1), 2 measures were established for 254 (66.5%), and 3 were established for 116 (30.4%). There was a notable prevalence of LTE orders on CPR, invasive measures, ICU hospitalization, parenteral nutrition, antibiotherapy, diagnostic tests and surgery (Table 3). The mean time from admission to the decision making on LTE was 78 ± 18 h (for ‘‘no CPR’’, the time was 85 ± 222 h). Terminal sedation was prescribed for 178 patients (47.6%), with a mean duration to death of 38 ± 45 h. Table 4 shows the association between LTE measures, the ‘‘no CPR’’ decision and the administration of palliative sedation on one hand and the degree of dementia,
Table 1 died.
3 Baseline characteristics of the 382 patients who
Category Age, years Sex, female Length of stay, days
85 ± 10 204 (53) 11 ± 13
Place of residence Institutionalized Home Unspecified
136 (36) 222 (58) 24 (6)
Dementiaa No Mild Moderate Severe Unknown
147 35 56 109 36
(38) (9) (15) (29) (9)
Comorbidityb Absent Mild Moderate Severe
2 15 101 251
(1) (4) (27) (68)
Terminal patient Unexpected death Prescribed sedation Duration of sedation, h Time from admission to the ‘‘do not resuscitate’’ order, h Time from admission to the LTE, h Time from LTE to the prescription of sedation, h
193 (51) 39 (10) 178 (48) 38 ± 45 85 ± 222 78 ± 181 112 ± 225
The results are expressed as n (%) or as mean ± standard deviation. Abbreviations: CPR, cardiopulmonary resuscitation; LTE, limitation of therapeutic effort. a According to the Clinical Dementia Rating criteria. b According to the age-adjusted Charlson index.
institutionalization, the Charlson index, sex and age on the other. A statistically significant association was observed only between the ‘‘no CPR’’ order and the presence of moderate to severe dementia (OR 1.73; 95% CI 1.03---2.91; chi-squared, 4.33; p = .03).
Discussion Our study shows that more than 80% of the patients who died in internal medicine hospital areas received some measure of LTE. In general, these patients were elderly, with high comorbidity, cognitive impairment and terminal illness. Occasionally, this decision is poorly reflected in the medical history, with expressions such as ‘‘do not perform aggressive measures’’. In our setting, few studies have addressed LTE in internal medicine departments. The study by Solis et al. focused on ‘‘do not resuscitate’’ orders in a district hospital.14 The population’s baseline characteristics were similar to those of the present study, except for a higher proportion of patients with cancer. However, the documentation of ‘‘do not
4
R. García Caballero et al. Table 2
Causes of hospitalization and death of the included patients.
Reasons for admission
n (%)
Respiratory infection Heart failure Urinary tract infection Other infections Acute stroke Renal failure Hepatobiliary disease Ischemic heart disease Pulmonary thromboembolism Others
Table 3 orders.
126 63 39 29 27 12 11 7 2 66
(33.0) (16.5) (10.2) (7.6) (7.1) (3.1) (2.9) (1.8) (0.50) (17.3)
Established limitation of therapeutic effort
Orders No CPR No aggressive measures No ICU admission No parenteral nutrition No antibiotherapy No diagnostic tests performed No surgical treatment No renal replacement therapy No blood product transfusion No chemotherapy or radiation therapy No ventilation/intubation Others
N (%)a 292 113 102 60 40 35 34 17 11 10
(76.4) (16.45) (14.85) (8.73) (5.82) (5.09) (4.95) (2.47) (1.60) (1.46)
5 (0.73) 1 (0.15)
95% CI 72.1---80.8 13.8---19.4 12.3---17.7 6.7---11.1 4.2---7.8 3.6---7.0 3.5---6.8 1.4---3.9 0.80---2.8 0.70---2.7 0.23---1.7 0.00---0.8
Abbreviations: CPR, cardiopulmonary resuscitation; ICU, intensive care unit. a A patient can have several prescribed orders.
resuscitate’’ orders was lower than that of our study (66% vs. 76%, respectively).14 Recently, data was reported from the 1457 patients included in the Care in their Last Days of Life of Patients in Internal Medicine Departments (UDVIMI) study, which is pending publication.21 The results were analogous to those of our study, although there were a number of differences, such as a lower percentage of expected deaths (62% vs. 90%), a longer time from admission to sedation (5 vs. 1.5 days) and shorter time to death (7 vs. 11 days). These differences could be explained by the different sample size and the duration of the patient enrollment. The UDVIMI study included the first 10 patients who died in the internal medicine departments of each center from December 1, 2015, the time of greatest healthcare burden during the winter period. Our study, however, conducted the enrollment over 6 months.22 According to the definition by the Spanish Society of Palliative Care, more than half of our study’s patients met the criteria for terminal illness at admission. However, the time elapsed to the LTE and the ‘‘do not resuscitate’’ order was long (78 and 85 h, respectively). Most terminal patients who are admitted to internal medicine departments are admitted for nononcologic diseases, which can lead to a delay in identifying the process as terminal.23,24 The delay in the
Causes of death Respiratory infection Heart failure Cancer Urinary tract infection Acute stroke Other infections Ischemic heart disease Pulmonary thromboembolism Others
n (%) 132 58 34 29 27 23 16 9 53
(34.5) (15.2) (8.9) (7.6) (7.1) (6.0) (4.2) (2.4) (13.8)
diagnosis and palliative treatment for these patients could be due to the difficulty in coming to a consensus on the end-of-life care with the patients and their relatives, a lack of training in communication skills and in palliative care (it is essential to identify terminal patients with nononcologic disease), work overload and fragmentation in the continuity of care.25 Although this study is focused on the hospital setting, the actions of primary care with regard to these patients are important. Occasionally, both the patients and their relatives prefer the hospital medium for the last days of life, while others prefer to die at home.26 Greater coordination between primary care and specialized care is therefore needed, as is promoting home palliative care.27 In our study, the presence of dementia to at least a moderate degree increased the probability of prescribing the ‘‘do not resuscitate’’ order by 73%. Unlike our study, other studies have shown an association between the LTE decision and patient origin (residence or home)28 and age.29 Other relevant issues not analyzed in our study include the sociocultural aspect30 and the individual who makes the decision regarding the LTE, be it the patient or the family.31 In terms of the study’s limitations, we should reiterate that this is an extension of a previous study,15 and therefore the dates for the data collection are different for each of the centers. We kept the same months in all years to avoid possible seasonal variability. Another limitation is the inclusion of only patients who died. It is possible that the patients with terminal illness who were discharged also had LTE measures, and their clinical characteristics might differ. Another limitation is the inherent biases of a retrospective study. Nevertheless, in terms of a possible lack of uniformity in the identification of the various variables collected, such as the LTE measures, the research team met several times to clarify the definition of the variables. It should also be noted that we could not obtain reliable information on the specific reason why the clinicians decided to adopt LTE measures. Lastly, we did not collect information on the administered drug combinations in sedation due to the lack of availability in most of the centers of the treatment sheets for the patients who died. In conclusion, the prescribing of LTE measures is very common among patients who die in internal medicine departments, the most prevalent of which are the ‘‘do not resuscitate’’ order and the unspecific expression ‘‘do not use aggressive measures’’. We need better training in
Limitation of therapeutic effort in hospitalized of internal medicine Table 4
5
Associations between limitation of therapeutic effort measures and patient characteristics.
Associations
OR
95% CI
2
p
Between LTE order and. . . Degree of dementiaa Institutionalization Charlson indexb Sex Age
0.97 0.77 1.18 1.00 Mann---Whitney U test
0.55---1.72 0.43---1.36 0.66---2.09 0.58---1.74
0.008 0.79 0.33 0.001
.93 .37 .56 .97 .97
Associations
OR
95% CI
2
p
Between no RCP order and. . . Degree of dementiaa Institutionalization Charlson indexb Sex Age
1.73 0.96 0.90 1.16 Mann---Whitney U test
1.03---2.91 0.57---1.60 0.55---1.47 0.72---1.88
4.33 0.02 0.17 0.40
.03 .88 .68 .52 .29
Associations
OR
95% CI
2
p
Between palliative sedation and. . . Degree of dementiaa Institutionalization Charlson indexb Sex Age
1.29 1.39 0.92 1.37 Mann---Whitney U test
0.83---1.99 0.90---2.14 0.60---1.41 0.91---2.06
1.35 2.23 0.13 2.32
.24 .13 .71 .12 .08
Abbreviations: CI, confidence interval; CPR, resuscitation cardiopulmonary; LTE, limitation of therapeutic effort; OR, odds ratio. a Absence or mild and moderate/severe dementia. b Age-adjusted Charlson index categorized according to the median of the sample (8 points).
palliative care, bioethics and communication skills in patients’ end-of-life period.
Conflict of interests The authors declare that they have no conflicts of interest.
Acknowledgements The authors would like to thank Rodrigo Alonso and Maria Manuela Barrera for their help in collecting the data at the University Hospital Foundation Alcorcon.
References 1. Herreros B, Moreno-Milán B, Pacho-Jiménez E, Real-de Asúa D, Roa-Castellanos RA, Valenti E. Terminología en bioética clínica. Rev Med Inst Mex Seguro Soc. 2015;53:750---61. 2. Martino Alba R, Monleón Luque M. Limitar el esfuerzo terapéutico: un término poco apropiado. Med Pal. 2009;16:9---10. 3. Newschaffer CJ, Bush TL, Penberthy LT. Comorbidity measurement in elderly female breast cancer patients with administrative and medical records data. J Clin Epidemiol. 1997;50:725---33. 4. Park SH, Lee HS. Assessing predictive validity of pressure Ulcer risk scales --- a systematic review and meta-analysis. Iran J Public Health. 2016;45:122---33. 5. Trujillo-Cari˜ no AL, Allende-Pérez S, Verástegui-Avilés E. Utilidad del índice pronóstico paliativo (PPI) en pacientes con cáncer. GAMO. 2013;12:234---9.
6. Cuadras Lacasa F, Alcaraz Benavides M, Llort Mateu M, Madriles Basaga˜ nas MS, Martín Ramírez C, Mesalles Sin M, et al. Índice de Karnofsky para medir la calidad de vida. Rev Enferm. 1998;21:18---20. 7. Esteban A, Gordo F, Solsona JF, Alía I, Caballero J, Bouza C, et al. Withdrawing and withholding life support in the intensive care unit: a Spanish prospective multi-centre observational study. Intens Care Med. 2001;27:1744---9. 8. Faull C, Rowe Haynes C, Oliver D. Issues for palliative medicine doctors surrounding the withdrawal of non-invasive ventilation at the request of a patient with motor neurone disease: a scoping study. BMJ Support Palliat Care. 2014;4:43---9. 9. Lim RB. End-of-life care in patients with advanced lung cancer. Ther Adv Respir Dis. 2016;10:455---67. 10. Arcand M. End-of-life issues in advanced dementia: Part 1: Goals of care, decision-making process, and family education. Can Fam Physician. 2015;61:330---4. 11. Vernaz S, Casanova L, Blanc F, Lebel S, Ughetto F, Paut O. To maintain or to withdraw life support? Variations on the methods of ending life in a pediatric intensive care unit over a period of 6 years. Ann Fr Anesth Reanim. 2014;33:400---4. 12. Castillo Rueda A, de Portugal Álvarez J. Proyecto técnico de gestión y funcionamiento de la unidad asistencial de Medicina Interna. An Med Interna. 2004;21:31---8. 13. Novillo A, Ladenheim R, Galante M, Isola IM, Musi ME, Naguel V, et al. Limitation of life-sustaining treatment. A prospective study in a clinical ward. Medicina (B Aires). 2008;68:437---41. 14. Solís-García del Pozo J, Gómez-Pérez I. The application of do not resuscitate orders and withholding treatment in patients admitted to internal medicine in a first level hospital. Rev Calid Asist. 2013;28:50---5. 15. García Caballero R, Herreros B, Real de Asúa D, Alonso R, Barrera MM, Castilla V. Limitación del esfuerzo terapéutico en
6
16.
17.
18. 19. 20.
21.
22. 23.
R. García Caballero et al. pacientes hospitalizados en el servicio de medicina interna. Rev Calid Asist. 2016;31:70---5. Asociación médica mundial. 64.a Asamblea General. Declaración de Helsinki de la AAM. Principios éticos para las investigaciones médicas en seres humanos. Fortaleza. Brasil. Octubre 2013. Available from: http://www.wma.net/ es/30publications/10policies/b3/ [accessed 19.01.17]. Ley 41/2002, de 14 de noviembre, básica reguladora de la autonomía del paciente y de derechos y obligaciones en materia de información y documentación clínica. BOE. n.◦ 274; 15/11/2002:22188. Morris, John C. The clinical dementia rating (CDR): current version and scoring rules. Neurology. 1993;43:2412---4. Charlson M, Szatrowski TP, Peterson J. Validation of a combined comorbidity index. J Clin Epidemiol. 1994;47:1245---51. Navarro Sanz JR. Cuidados paliativos no oncológicos: enfermedad terminal y factores pronósticos. [Guía Médica]. Madrid: SECPAL. Available from: http://www.secpal.com/CUIDADOSPALIATIVOS-NO-ONCOLOGICOS-ENFERMEDAD-TERMINAL-YFACTORES-PRONOSTICOS [accessed 19.01.17]. Isasi de Isasmendi S, Rubio Gómez M, Díez Manglano J. Uso de la sedación paliativa en medicina interna (estudio UDVIMI). Comunicación presentada al XXXVII Congreso Nacional nola de Medicina Interna. Zaragoza. de la Sociedad Espa˜ Espa˜ na; Noviembre de 2016. Available from: http://www. revclinesp.es/es/congresos/xxxvii-congreso-nacional-sociedadespanola/44/sesion/paciente-pluripatologico-edad-avanzada/ 2809/uso-de-la-sedación-paliativa/30591/ [accessed 21.05. 17]. Díez-Manglano J. La pluripatología, un reto para los sistemas sanitarios. Rev Clin Esp. 2017;217:207. Bernabeu-Wittel M, García-Morillo S, González-Becerra C, Ollero M, Fernández A, Cuello-Contreras JA. Impacto de los
24.
25.
26. 27.
28.
29.
30.
31.
cuidados paliativos y perfil clínico del paciente con enfermedad terminal en un área de Medicina Interna. Rev Clin Esp. 2006;206:178---81. Le Conte P, Riochet D, Batard E, Volteau C, Giraudeau B, Arnaudet I, et al. Death in emergency departments: a multicenter cross-sectional survey with analysis of withholding and withdrawing life support. Intensive Care Med. 2010;36: 765---7. Pérez Rueda M [tesis doctoral] Conocimiento y actitudes ante las instrucciones previas de pacientes y profesionales sanitarios en la Comunidad de Madrid. Madrid: Departamento de Toxicología y Legislación sanitaria, Universidad Complutense de Madrid; 2015. Gervás J. Morir en casa con dignidad. Una posibilidad, si hay apoyo y cuidados de calidad. Gac Med Bilbao. 2011;108:3---6. Garrido Sanjuán JA. El internista y las competencias en el área na Interna (SOGAMI). La de bioética. Sociedad Galega de Medici˜ na. medicina interna como modelo de práctica clínica. A Coru˜ 2008:231---45. Parsons C, McCorry N, Murphy K, Byrne S, O’Sullivan D, O’Mahony D, et al. Assessment of factors that influence physician decision making regarding medication use in patients with dementia at the end of life. Int J Geriatr Psychiatry. 2014;29:281---90. Gillick MR, Mendes ML. Medical care in old age: what do nurses in long-term care consider appropriate? J Am Geriatr Soc. 1996;44:1322---5. Lu CY, Johantgen M. Factors associated with treatment restriction orders and hospice in older nursing home residents. J Clin Nurs. 2011;20:377---87. Torian LV, Davidson EJ, Fillit HM, Fulop G, Sell LL. Decisions for and against resuscitation in an acute geriatric medicine unit serving the frail elderly. Arch Intern Med. 1992;152:561---5.
Revista Clínica Española www.elsevier.es/rce
ORIGINAL ARTICLE
Limitation of therapeutic effort in patients hospitalized in departments of internal medicine夽,夽夽 R. García Caballero a,b,c,∗ , B. Herreros a,b,c , D. Real de Asúa a,c,d , S. Gámez e , G. Vega f , L. García Olmos g a
Grupo de Trabajo de Bioética de la Sociedad Espa˜ nola de Medicina Interna, Madrid, Spain Unidad de Medicina Interna, Hospital Universitario Fundación Alcorcón, Alcorcón, Madrid, Spain c Instituto de Ética Clínica Francisco Vallés, Universidad Europea, Madrid, Spain d Servicio de Medicina Interna, Hospital Universitario de La Princesa, Madrid, Spain e Hospital Universitario Infanta Cristina, Parla, Madrid, Spain f Hospital Universitario Quironsalud, Pozuelo de Alarcón, Madrid, Spain g Unidad docente multiprofesional de Atención Familiar y Comunitaria, Sureste, Madrid, Spain b
Received 26 June 2017; accepted 3 October 2017 Available online 2 December 2017
KEYWORDS Limitation of therapeutic effort; No cardiopulmonary resuscitation; Decision making; Internal medicine
Abstract Introduction: There is little information on the limitation of therapeutic effort (LTE) in patients admitted to hospital internal medicine units. Objectives: To describe the indicated LTE regimens in the departments of internal medicine and the characteristics of the patients who undergo them. Patients and methods: An observational, descriptive retrospective study was conducted on 4 hospitals of the Community of Madrid. The study collected demographic and comorbidity data and the LTE orders prescribed for all patients who died during a period of 6 months. Results: The study included 382 patients with a mean age of 85 ± 10 years; 204 were women (53.4%) and 222 (58.1%) came from their homes. Some 51.1% of the patients were terminal, 43.2% had moderate to severe dementia, and 95.5% presented at least moderate comorbidity. Some type of LTE was performed in 318 patients (83.7%); the most common orders were ‘‘No cardiopulmonary resuscitation’’ (292 patients, 76.4%; 95% CI 72.1---80.8), ‘‘Do not use aggressive measures’’ (113 patients, 16.4%; 95% CI 13.7---19.4) and ‘‘Do not transfer to an intensive care unit’’ (102 cases, 14.8%; 95% CI 12.3---17.7). Some type of LTE was performed in 318 patients
夽 Please cite this article as: García Caballero R, Herreros B, Real de Asúa D, Gámez S, Vega G, García Olmos L. Limitación del esfuerzo terapéutico en pacientes hospitalizados en servicios de medicina interna. Rev Clin Esp. 2018;218:1---6. 夽夽 The study is part of doctoral thesis project by Rebeca García Caballero. ∗ Corresponding author. E-mail address: [email protected] (R. García Caballero).
2254-8874/© 2017 Elsevier Espa˜ na, S.L.U. and Sociedad Espa˜ nola de Medicina Interna (SEMI). All rights reserved.
2
R. García Caballero et al. (83.7%); the most common orders were ‘‘No cardiopulmonary resuscitation’’ (292 patients, 76.4%; 95% CI 72.1---80.8), ‘‘Do not use aggressive measures’’ (113 patients, 16.4%; 95% CI 13.7---19.4) and ‘‘Do not transfer to an intensive care unit’’ (102 cases, 14.8%; 95% CI 12.3---17.7). Conclusions: LTE is common among patients who die in Internal Medicine. The most widely used regimens were ‘‘No CPR’’ and the unspecific statement ‘‘Do not use aggressive measures’’. The patients were elderly and had significant comorbidity, terminal illness and advanced dementia. © 2017 Elsevier Espa˜ na, S.L.U. and Sociedad Espa˜ nola de Medicina Interna (SEMI). All rights reserved.
PALABRAS CLAVE Limitación del esfuerzo terapéutico; Órdenes de no reanimación cardiopulmonar; Toma de decisiones; Medicina interna
Limitación del esfuerzo terapéutico en pacientes hospitalizados en servicios de medicina interna Resumen Introducción: Existe escasa información sobre la limitación del esfuerzo terapéutico (LET) en pacientes ingresados en unidades de hospitalización de medicina interna. Objetivos: Describir las pautas de LET indicadas en los servicios de medicina interna y las características de los pacientes que las reciben. Pacientes y métodos: Estudio observacional descriptivo y retrospectivo de 4 hospitales de la Comunidad de Madrid. Se recogieron datos demográficos, de comorbilidad y las órdenes de LET pautadas en todos los pacientes fallecidos en un periodo de 6 meses. Resultados: Se incluyeron 382 pacientes cuya edad media fue de 85 ± 10 a˜ nos; 204 eran mujeres (53,4%) y 222 (58,1%) procedían de su domicilio. El 51,1% eran enfermos terminales, el 43,2% tenían demencia moderada/grave y el 95,5% presentaban comorbilidad al menos moderada. En 318 pacientes (83,7%) se realizó algún tipo de LET, siendo las más frecuentes las órdenes de «no reanimación cardiopulmonar» (292 enfermos, 76,4%; IC 95%: 72,1-80,8), «no usar medidas agresivas» (113 pacientes, 16,4%; IC 95%: 13,7-19,4) y «no ingresar en unidad de cuidados intensivos» (102 casos, 14,8%; IC 95%: 12,3-17,7). Conclusiones: La LET es muy frecuente en los pacientes que fallecen en medicina interna. Las pautas más utilizadas son «no reanimación cardiopulmonar» y la expresión poco concreta de «no usar medidas agresivas». Los pacientes son de edad avanzada, con importante comorbilidad, enfermedad terminal y demencia avanzada. © 2017 Elsevier Espa˜ na, S.L.U. y Sociedad Espa˜ nola de Medicina Interna (SEMI). Todos los derechos reservados.
Background Numerous expressions have been used to refer to the limitation of therapeutic effort (LTE), but none of them accurately explain the concept. A widely accepted definition is ‘‘not applying measures disproportionate to the therapeutic goal (mainly healing, improving or relieving) to patients with poor vital prognoses and a poor quality of life’’.1,2 A disproportionate measure is understood as an imbalance between costs-charges and benefits for the patient. LTE attempts to prevent unnecessary suffering in patients in their final phase of life.1 To consider a patient a candidate for LTE measures, various scales have been implemented that include factors such as quality of life, comorbidity, risk and vital prognosis.3---6 LTE has been mainly evaluated in other health departments such as intensive care,7 palliative care,8 oncology,9 geriatric medicine10 and pediatrics.11 The departments of internal medicine treat a significant percentage of hospitalized patients, many of them in their final period of life.12 However, there is very little consistent information on the LTE measures to be applied to these patients,13,14 whose demographic and clinical characteristics
differ significantly from those of patients hospitalized in other departments. This study describes the LTE guidelines indicated in internal medicine departments and the demographic and clinical characteristics of the patients who receive these guidelines. This study is the continuation and extension of a previous project (Withholding and withdrawing treatment in patients admitted in an Internal Medicine ward), where we analyzed the LTE in a single hospital.15
Patients and methods This was an observational, descriptive, retrospective multicenter study on 4 hospitals of the Community of Madrid, which provide care to urban and rural populations. We included all those patients who died during their hospitalization; there were no exclusion criteria. The study lasted 6 months and was conducted according to good clinical practice criteria and the Declaration of Helsinki.16,17 We obtained the following variables from the medical records: age, sex, hospital stay, patient origin (institution/residence vs. home), cognitive impairment (according to the Clinical Dementia Rating scale),18 comorbidity
Limitation of therapeutic effort in hospitalized of internal medicine (age-adjusted Charlson index: absence of comorbidity [1---2 points], mild [3---4 points], moderate [5---6 points] and severe comorbidity [>7 points]19 ), terminal status criteria at admission (according to the Spanish Society of Palliative Care20 ), reason for hospitalization and cause of death. The analyzed LTE measures were no cardiopulmonary resuscitation (CPR) order, hospitalization in the intensive care unit, use of renal replacement therapy, blood-product transfusion, implementation of diagnostic tests, surgery, parenteral nutrition, chemotherapy or radiation therapy, mechanical ventilation or intubation and the nonimplementation of ‘‘aggressive measures’’. We recorded only those orders that explicitly appeared in the medical progress notes. We also recorded the time elapsed from admission to the LTE order, as well as the palliative sedation measures1 and the duration of sedation. The data were analyzed with SPSS, version 17.0 (IBM Corp., Armonk, NY). We calculated the 95% confidence intervals of the primary endpoint (LTE regimen). The qualitative variables are presented using frequency tables (percentages), and the quantitative variables are presented as mean and standard deviation. We analyzed the associations between the LTE orders, abstention from CPR and palliative sedation on one hand and the following variables on the other: degree of dementia, institutionalization, Charlson index, sex and age. To perform the analysis, we categorized the Charlson index using the median as the cutoff (8 points). The degree of dementia was grouped into 2 categories: absent/mild (grade 0---1 on the Clinical Dementia Rating scale) and moderate/severe dementia (grade 2---3). We calculated the odds ratio (OR), its confidence intervals and the significance level using the chi-squared test for the contrast of hypothesis.
Results The study analyzed 382 patients with a mean age of 85 ± 10 years; 204 were women (53.4%) (Table 1), and the majority (222, 58.1%) came from their homes. The mean hospital stay was 11 ± 13 days. Some 43.2% of the patients had moderate-severe dementia, 95.5% had at least moderate comorbidity, and 51.1% met terminal status criteria. There were 39 (10.3%) unexpected deaths, understood as those for whom there were no comments indicative of imminent death in the medical and nursing notes in the days prior to the death. Other general demographic and clinical characteristics of the sample are shown in Table 1. The causes of hospitalization and death are indicated in Table 2. An LTE measure was established for 318 patients (83.7%; 95% CI 79.3---87.1), 2 measures were established for 254 (66.5%), and 3 were established for 116 (30.4%). There was a notable prevalence of LTE orders on CPR, invasive measures, ICU hospitalization, parenteral nutrition, antibiotherapy, diagnostic tests and surgery (Table 3). The mean time from admission to the decision making on LTE was 78 ± 18 h (for ‘‘no CPR’’, the time was 85 ± 222 h). Terminal sedation was prescribed for 178 patients (47.6%), with a mean duration to death of 38 ± 45 h. Table 4 shows the association between LTE measures, the ‘‘no CPR’’ decision and the administration of palliative sedation on one hand and the degree of dementia,
Table 1 died.
3 Baseline characteristics of the 382 patients who
Category Age, years Sex, female Length of stay, days
85 ± 10 204 (53) 11 ± 13
Place of residence Institutionalized Home Unspecified
136 (36) 222 (58) 24 (6)
Dementiaa No Mild Moderate Severe Unknown
147 35 56 109 36
(38) (9) (15) (29) (9)
Comorbidityb Absent Mild Moderate Severe
2 15 101 251
(1) (4) (27) (68)
Terminal patient Unexpected death Prescribed sedation Duration of sedation, h Time from admission to the ‘‘do not resuscitate’’ order, h Time from admission to the LTE, h Time from LTE to the prescription of sedation, h
193 (51) 39 (10) 178 (48) 38 ± 45 85 ± 222 78 ± 181 112 ± 225
The results are expressed as n (%) or as mean ± standard deviation. Abbreviations: CPR, cardiopulmonary resuscitation; LTE, limitation of therapeutic effort. a According to the Clinical Dementia Rating criteria. b According to the age-adjusted Charlson index.
institutionalization, the Charlson index, sex and age on the other. A statistically significant association was observed only between the ‘‘no CPR’’ order and the presence of moderate to severe dementia (OR 1.73; 95% CI 1.03---2.91; chi-squared, 4.33; p = .03).
Discussion Our study shows that more than 80% of the patients who died in internal medicine hospital areas received some measure of LTE. In general, these patients were elderly, with high comorbidity, cognitive impairment and terminal illness. Occasionally, this decision is poorly reflected in the medical history, with expressions such as ‘‘do not perform aggressive measures’’. In our setting, few studies have addressed LTE in internal medicine departments. The study by Solis et al. focused on ‘‘do not resuscitate’’ orders in a district hospital.14 The population’s baseline characteristics were similar to those of the present study, except for a higher proportion of patients with cancer. However, the documentation of ‘‘do not
4
R. García Caballero et al. Table 2
Causes of hospitalization and death of the included patients.
Reasons for admission
n (%)
Respiratory infection Heart failure Urinary tract infection Other infections Acute stroke Renal failure Hepatobiliary disease Ischemic heart disease Pulmonary thromboembolism Others
Table 3 orders.
126 63 39 29 27 12 11 7 2 66
(33.0) (16.5) (10.2) (7.6) (7.1) (3.1) (2.9) (1.8) (0.50) (17.3)
Established limitation of therapeutic effort
Orders No CPR No aggressive measures No ICU admission No parenteral nutrition No antibiotherapy No diagnostic tests performed No surgical treatment No renal replacement therapy No blood product transfusion No chemotherapy or radiation therapy No ventilation/intubation Others
N (%)a 292 113 102 60 40 35 34 17 11 10
(76.4) (16.45) (14.85) (8.73) (5.82) (5.09) (4.95) (2.47) (1.60) (1.46)
5 (0.73) 1 (0.15)
95% CI 72.1---80.8 13.8---19.4 12.3---17.7 6.7---11.1 4.2---7.8 3.6---7.0 3.5---6.8 1.4---3.9 0.80---2.8 0.70---2.7 0.23---1.7 0.00---0.8
Abbreviations: CPR, cardiopulmonary resuscitation; ICU, intensive care unit. a A patient can have several prescribed orders.
resuscitate’’ orders was lower than that of our study (66% vs. 76%, respectively).14 Recently, data was reported from the 1457 patients included in the Care in their Last Days of Life of Patients in Internal Medicine Departments (UDVIMI) study, which is pending publication.21 The results were analogous to those of our study, although there were a number of differences, such as a lower percentage of expected deaths (62% vs. 90%), a longer time from admission to sedation (5 vs. 1.5 days) and shorter time to death (7 vs. 11 days). These differences could be explained by the different sample size and the duration of the patient enrollment. The UDVIMI study included the first 10 patients who died in the internal medicine departments of each center from December 1, 2015, the time of greatest healthcare burden during the winter period. Our study, however, conducted the enrollment over 6 months.22 According to the definition by the Spanish Society of Palliative Care, more than half of our study’s patients met the criteria for terminal illness at admission. However, the time elapsed to the LTE and the ‘‘do not resuscitate’’ order was long (78 and 85 h, respectively). Most terminal patients who are admitted to internal medicine departments are admitted for nononcologic diseases, which can lead to a delay in identifying the process as terminal.23,24 The delay in the
Causes of death Respiratory infection Heart failure Cancer Urinary tract infection Acute stroke Other infections Ischemic heart disease Pulmonary thromboembolism Others
n (%) 132 58 34 29 27 23 16 9 53
(34.5) (15.2) (8.9) (7.6) (7.1) (6.0) (4.2) (2.4) (13.8)
diagnosis and palliative treatment for these patients could be due to the difficulty in coming to a consensus on the end-of-life care with the patients and their relatives, a lack of training in communication skills and in palliative care (it is essential to identify terminal patients with nononcologic disease), work overload and fragmentation in the continuity of care.25 Although this study is focused on the hospital setting, the actions of primary care with regard to these patients are important. Occasionally, both the patients and their relatives prefer the hospital medium for the last days of life, while others prefer to die at home.26 Greater coordination between primary care and specialized care is therefore needed, as is promoting home palliative care.27 In our study, the presence of dementia to at least a moderate degree increased the probability of prescribing the ‘‘do not resuscitate’’ order by 73%. Unlike our study, other studies have shown an association between the LTE decision and patient origin (residence or home)28 and age.29 Other relevant issues not analyzed in our study include the sociocultural aspect30 and the individual who makes the decision regarding the LTE, be it the patient or the family.31 In terms of the study’s limitations, we should reiterate that this is an extension of a previous study,15 and therefore the dates for the data collection are different for each of the centers. We kept the same months in all years to avoid possible seasonal variability. Another limitation is the inclusion of only patients who died. It is possible that the patients with terminal illness who were discharged also had LTE measures, and their clinical characteristics might differ. Another limitation is the inherent biases of a retrospective study. Nevertheless, in terms of a possible lack of uniformity in the identification of the various variables collected, such as the LTE measures, the research team met several times to clarify the definition of the variables. It should also be noted that we could not obtain reliable information on the specific reason why the clinicians decided to adopt LTE measures. Lastly, we did not collect information on the administered drug combinations in sedation due to the lack of availability in most of the centers of the treatment sheets for the patients who died. In conclusion, the prescribing of LTE measures is very common among patients who die in internal medicine departments, the most prevalent of which are the ‘‘do not resuscitate’’ order and the unspecific expression ‘‘do not use aggressive measures’’. We need better training in
Limitation of therapeutic effort in hospitalized of internal medicine Table 4
5
Associations between limitation of therapeutic effort measures and patient characteristics.
Associations
OR
95% CI
2
p
Between LTE order and. . . Degree of dementiaa Institutionalization Charlson indexb Sex Age
0.97 0.77 1.18 1.00 Mann---Whitney U test
0.55---1.72 0.43---1.36 0.66---2.09 0.58---1.74
0.008 0.79 0.33 0.001
.93 .37 .56 .97 .97
Associations
OR
95% CI
2
p
Between no RCP order and. . . Degree of dementiaa Institutionalization Charlson indexb Sex Age
1.73 0.96 0.90 1.16 Mann---Whitney U test
1.03---2.91 0.57---1.60 0.55---1.47 0.72---1.88
4.33 0.02 0.17 0.40
.03 .88 .68 .52 .29
Associations
OR
95% CI
2
p
Between palliative sedation and. . . Degree of dementiaa Institutionalization Charlson indexb Sex Age
1.29 1.39 0.92 1.37 Mann---Whitney U test
0.83---1.99 0.90---2.14 0.60---1.41 0.91---2.06
1.35 2.23 0.13 2.32
.24 .13 .71 .12 .08
Abbreviations: CI, confidence interval; CPR, resuscitation cardiopulmonary; LTE, limitation of therapeutic effort; OR, odds ratio. a Absence or mild and moderate/severe dementia. b Age-adjusted Charlson index categorized according to the median of the sample (8 points).
palliative care, bioethics and communication skills in patients’ end-of-life period.
Conflict of interests The authors declare that they have no conflicts of interest.
Acknowledgements The authors would like to thank Rodrigo Alonso and Maria Manuela Barrera for their help in collecting the data at the University Hospital Foundation Alcorcon.
References 1. Herreros B, Moreno-Milán B, Pacho-Jiménez E, Real-de Asúa D, Roa-Castellanos RA, Valenti E. Terminología en bioética clínica. Rev Med Inst Mex Seguro Soc. 2015;53:750---61. 2. Martino Alba R, Monleón Luque M. Limitar el esfuerzo terapéutico: un término poco apropiado. Med Pal. 2009;16:9---10. 3. Newschaffer CJ, Bush TL, Penberthy LT. Comorbidity measurement in elderly female breast cancer patients with administrative and medical records data. J Clin Epidemiol. 1997;50:725---33. 4. Park SH, Lee HS. Assessing predictive validity of pressure Ulcer risk scales --- a systematic review and meta-analysis. Iran J Public Health. 2016;45:122---33. 5. Trujillo-Cari˜ no AL, Allende-Pérez S, Verástegui-Avilés E. Utilidad del índice pronóstico paliativo (PPI) en pacientes con cáncer. GAMO. 2013;12:234---9.
6. Cuadras Lacasa F, Alcaraz Benavides M, Llort Mateu M, Madriles Basaga˜ nas MS, Martín Ramírez C, Mesalles Sin M, et al. Índice de Karnofsky para medir la calidad de vida. Rev Enferm. 1998;21:18---20. 7. Esteban A, Gordo F, Solsona JF, Alía I, Caballero J, Bouza C, et al. Withdrawing and withholding life support in the intensive care unit: a Spanish prospective multi-centre observational study. Intens Care Med. 2001;27:1744---9. 8. Faull C, Rowe Haynes C, Oliver D. Issues for palliative medicine doctors surrounding the withdrawal of non-invasive ventilation at the request of a patient with motor neurone disease: a scoping study. BMJ Support Palliat Care. 2014;4:43---9. 9. Lim RB. End-of-life care in patients with advanced lung cancer. Ther Adv Respir Dis. 2016;10:455---67. 10. Arcand M. End-of-life issues in advanced dementia: Part 1: Goals of care, decision-making process, and family education. Can Fam Physician. 2015;61:330---4. 11. Vernaz S, Casanova L, Blanc F, Lebel S, Ughetto F, Paut O. To maintain or to withdraw life support? Variations on the methods of ending life in a pediatric intensive care unit over a period of 6 years. Ann Fr Anesth Reanim. 2014;33:400---4. 12. Castillo Rueda A, de Portugal Álvarez J. Proyecto técnico de gestión y funcionamiento de la unidad asistencial de Medicina Interna. An Med Interna. 2004;21:31---8. 13. Novillo A, Ladenheim R, Galante M, Isola IM, Musi ME, Naguel V, et al. Limitation of life-sustaining treatment. A prospective study in a clinical ward. Medicina (B Aires). 2008;68:437---41. 14. Solís-García del Pozo J, Gómez-Pérez I. The application of do not resuscitate orders and withholding treatment in patients admitted to internal medicine in a first level hospital. Rev Calid Asist. 2013;28:50---5. 15. García Caballero R, Herreros B, Real de Asúa D, Alonso R, Barrera MM, Castilla V. Limitación del esfuerzo terapéutico en
6
16.
17.
18. 19. 20.
21.
22. 23.
R. García Caballero et al. pacientes hospitalizados en el servicio de medicina interna. Rev Calid Asist. 2016;31:70---5. Asociación médica mundial. 64.a Asamblea General. Declaración de Helsinki de la AAM. Principios éticos para las investigaciones médicas en seres humanos. Fortaleza. Brasil. Octubre 2013. Available from: http://www.wma.net/ es/30publications/10policies/b3/ [accessed 19.01.17]. Ley 41/2002, de 14 de noviembre, básica reguladora de la autonomía del paciente y de derechos y obligaciones en materia de información y documentación clínica. BOE. n.◦ 274; 15/11/2002:22188. Morris, John C. The clinical dementia rating (CDR): current version and scoring rules. Neurology. 1993;43:2412---4. Charlson M, Szatrowski TP, Peterson J. Validation of a combined comorbidity index. J Clin Epidemiol. 1994;47:1245---51. Navarro Sanz JR. Cuidados paliativos no oncológicos: enfermedad terminal y factores pronósticos. [Guía Médica]. Madrid: SECPAL. Available from: http://www.secpal.com/CUIDADOSPALIATIVOS-NO-ONCOLOGICOS-ENFERMEDAD-TERMINAL-YFACTORES-PRONOSTICOS [accessed 19.01.17]. Isasi de Isasmendi S, Rubio Gómez M, Díez Manglano J. Uso de la sedación paliativa en medicina interna (estudio UDVIMI). Comunicación presentada al XXXVII Congreso Nacional nola de Medicina Interna. Zaragoza. de la Sociedad Espa˜ Espa˜ na; Noviembre de 2016. Available from: http://www. revclinesp.es/es/congresos/xxxvii-congreso-nacional-sociedadespanola/44/sesion/paciente-pluripatologico-edad-avanzada/ 2809/uso-de-la-sedación-paliativa/30591/ [accessed 21.05. 17]. Díez-Manglano J. La pluripatología, un reto para los sistemas sanitarios. Rev Clin Esp. 2017;217:207. Bernabeu-Wittel M, García-Morillo S, González-Becerra C, Ollero M, Fernández A, Cuello-Contreras JA. Impacto de los
24.
25.
26. 27.
28.
29.
30.
31.
cuidados paliativos y perfil clínico del paciente con enfermedad terminal en un área de Medicina Interna. Rev Clin Esp. 2006;206:178---81. Le Conte P, Riochet D, Batard E, Volteau C, Giraudeau B, Arnaudet I, et al. Death in emergency departments: a multicenter cross-sectional survey with analysis of withholding and withdrawing life support. Intensive Care Med. 2010;36: 765---7. Pérez Rueda M [tesis doctoral] Conocimiento y actitudes ante las instrucciones previas de pacientes y profesionales sanitarios en la Comunidad de Madrid. Madrid: Departamento de Toxicología y Legislación sanitaria, Universidad Complutense de Madrid; 2015. Gervás J. Morir en casa con dignidad. Una posibilidad, si hay apoyo y cuidados de calidad. Gac Med Bilbao. 2011;108:3---6. Garrido Sanjuán JA. El internista y las competencias en el área na Interna (SOGAMI). La de bioética. Sociedad Galega de Medici˜ na. medicina interna como modelo de práctica clínica. A Coru˜ 2008:231---45. Parsons C, McCorry N, Murphy K, Byrne S, O’Sullivan D, O’Mahony D, et al. Assessment of factors that influence physician decision making regarding medication use in patients with dementia at the end of life. Int J Geriatr Psychiatry. 2014;29:281---90. Gillick MR, Mendes ML. Medical care in old age: what do nurses in long-term care consider appropriate? J Am Geriatr Soc. 1996;44:1322---5. Lu CY, Johantgen M. Factors associated with treatment restriction orders and hospice in older nursing home residents. J Clin Nurs. 2011;20:377---87. Torian LV, Davidson EJ, Fillit HM, Fulop G, Sell LL. Decisions for and against resuscitation in an acute geriatric medicine unit serving the frail elderly. Arch Intern Med. 1992;152:561---5.