Abstract
Objective:
To develop an enhanced understanding of factors that influence providers’ views about end-of-life care, we examined the contributions of provider, hospital, and country to variability in agreement with consensus statements about end-of-life care.
Design and Setting:
Data were drawn from a survey of providers’ views on principles of end-of-life care obtained during the consensus process for the Worldwide End-of-Life Practice for Patients in Intensive Care Units (WELPICUS) study.
Subjects:
Participants in WELPICUS included physicians, nurses and other providers. Our sample included 1068 providers from 178 hospitals and 31 countries.
Interventions:
None
Measurements and Main Results:
We examined views on cardiopulmonary resuscitation and withholding/withdrawing life-sustaining treatments, using a 3-level linear mixed model of responses from providers within hospitals within countries. Of 1,068 providers from 178 hospitals and 31 countries, 1% strongly disagreed, 7% disagreed, 11% were neutral, 44% agreed, and 36% strongly agreed with declining to offer cardiopulmonary resuscitation when not indicated. Of the total variability in those responses, 98%, 0%, and 2% was explained by differences among providers, hospitals, and countries, respectively. After accounting for provider characteristics and hospital size, the variance partition was similar. Results were similar for withholding/withdrawing life-sustaining treatments.
Conclusions:
Variability in agreement with consensus statements about end-of-life care is related primarily to differences among providers. Acknowledging the primary source of variability may facilitate efforts to achieve consensus and improve decision-making for critically ill patients and their family members at the end of life.
Keywords: End-of-life care, Intensive care unit, Cardiopulmonary resuscitation, Withholding life-sustaining treatment, Withdrawing life-sustaining treatment
Introduction
End-of-life care in the intensive care unit (ICU) varies widely, with differences in attitudes and practice patterns identified across individual providers, hospitals, and geographical regions (1–12). Given the inherent complexities involved in medical decision-making at the end of life, some degree of variability is expected. However, substantial variation in end-of-life care across providers, hospitals, or regions should be scrutinized, particularly when specific practice patterns may be associated with worse patient- and family-centered outcomes (13, 14).
There have been several attempts to achieve consensus on principles of end-of-life care for critically ill patients (13, 15, 16). The Worldwide End-of-Life Practice for Patients in Intensive Care Units (WELPICUS) study was the largest such effort. The WELPICUS study assessed provider views about 22 high-priority end-of-life issues (17). For each proposed statement about end-of-life care, participants provided responses using a 5-point scale, ranging from strongly disagree to strongly agree. Although consensus was achieved for most topics, there was variability in responses across providers. Our objective was to understand the factors that contributed to this variation. We identified 3 different levels which could potentially contribute to variability in agreement with consensus statements about end-of-life care: 1) individual respondents within a single hospital; 2) hospitals within a single country; and 3) countries. We assessed the contribution of each level to variance in providers’ agreement with consensus statements about end-of-life care. In addition, we examined associations between provider- and hospital-level predictors and agreement with these consensus statements. We hypothesized that most of the variability would exist at the provider level.
Methods
Study Design and Participants
This study was based on a 3-level design: individual respondents within hospitals within countries. Data were obtained from the WELPICUS study, which was designed to develop worldwide professional consensus for key end-of-life practices and published in 2014 (17). As a part of the WELPICUS study, critical care professional societies across the world worked together to identify participating centers, and a steering committee developed 35 definitions and 46 consensus statements for 22 issues related to end-of-life care. Consensus statements were posted in an electronic format, and participating providers were asked to document level of agreement with each item. Surveys were sent to 3,049 potential participants, with 1,283 initial responses from physicians (61%), nurses (30%), social workers (3%), ethicists (3%), attorneys (1%), and others (4%). Following the initial round of surveys, an additional 83 responses to the original questionnaire were received, providing a total sample of 1,366 participants from 192 hospitals and 32 countries that constituted the dataset from which our analyses were conducted. For the WELPICUS study, institutional review board approval was obtained (or determined to be exempt) for each center.
Measures
We focused on two areas addressed in WELPICUS: 1) declining to offer cardiopulmonary resuscitation when not indicated; and (2) withholding or withdrawing life-sustaining treatments (LST) with or without consent. Outcomes of interest were participants’ agreement with 5 consensus statements (Table 1). Providers documented their level of agreement with each statement, using a 5-point scale (strongly agree, agree, neutral, disagree, strongly disagree).
Table 1.
WELPICUS Consensus Statements.
| Topic | Consensus Statement |
|---|---|
| Declining to offer CPR when not indicated | If CPR is not indicated, this therapy should not be offered to patients or their surrogate decision-makers as if it were indicated. |
| Permissible to withhold LST | If a medical decision is made that a patient’s chances of surviving are extremely low or the patient under the present medical circumstances would not want continued life-sustaining treatment, life-sustaining treatment may be withheld. |
| Permissible to withhold LST without consent | Although life-sustaining treatment should generally be withheld only after obtaining informed consent of the patient and/or the surrogate decision-maker or family, there are circumstances when withholding life-sustaining treatment is permissible even though informed consent cannot be obtained (such as when the patient is not capable of decision-making and no family is available). |
| Permissible to withdraw LST | If a medical decision is made that a patient’s chances of surviving are extremely low or the patient under the present medical circumstances would not want continued life-sustaining treatment, life-sustaining treatment may be withdrawn. |
| Permissible to withdraw LST without consent | Although life-sustaining treatment should generally be withdrawn only after obtaining informed consent of the patient and/or the surrogate decision-maker or family, there are circumstances when withdrawing life-sustaining treatment is permissible even though informed consent cannot be obtained (such as when the patient is not capable of decision making and no family is available). |
We assessed 5 provider-level characteristics: age; sex; religion (Christian or non-Christian); profession (physician, nurse, other); and the frequency with which the provider was involved in decisions to withhold or withdraw LST (never/seldom, occasionally, frequently, regularly). One hospital-level characteristic was assessed: hospital size (<250 beds, 250–499, 500–999, 1000+).
Data Analysis
We included only providers with complete data for all outcomes and predictors. We treated the outcome as continuous: strongly disagree = 100 to strongly agree = 500 (we used a difference of 100 between categories to avoid decimals in results). Using linear mixed effects models, we assessed the proportion of the variability in the outcome explained by differences among providers, hospitals, and countries before and after the inclusion of fixed effects for provider- and hospital-level characteristics (18). The frequency of involvement in decisions to withhold or withdraw LST and hospital size was treated as continuous (with the difference between categories equal to 1). To evaluate the sensitivity of our results about the variance partition to treating the outcome as continuous, we also treated the outcome as binary (0=strongly disagree, disagree, or neutral; and 1=agree or strongly agree) and modeled it with a generalized linear mixed model (GLMM) with a logit link. Since the lowest level (provider) of variability in the GLMM is on the logistic scale and not directly comparable to variability at the other levels (hospital and country), we then used a linearization approach to obtain the variance partitions (19). Variance partition results based on this method were similar to those obtained using the linear mixed model (Supplementary Table 1). We fit the models in R with the lmer and glmer functions, and obtained 95% confidence intervals and p-values via the parametric bootstrap (18).
Results
Of the 1,366 providers from 192 hospitals and 32 countries with responses, there was complete data for 1068 providers (78%) from 178 hospitals and 31 countries (Table 2). The majority of providers were physicians (64%). Most providers cited occasional, frequent, or regular involvement in decisions to withhold or withdraw LST (76%). Most providers agreed or strongly agreed with each consensus statement (Table 3).
Table 2.
Characteristics of Respondents, Hospitals, and Countries
| Level | Number | Characteristic | N (%) or mean (SD) |
|---|---|---|---|
| Provider | 1,068 | Age, mean (SD) | 45 (9.3) |
| Female | 586 (55) | ||
| Religion | |||
| Christian | 651 (61) | ||
| Atheist/no religion | 275 (26) | ||
| Other | 142 (13) | ||
| Profession | |||
| Physician | 680 (64) | ||
| Nurse | 283 (27) | ||
| Othera | 105 (10) | ||
| Frequency of involvement in decisions to withhold/withdraw LST | |||
| Never/seldom | 252 (24) | ||
| Occasionally | 310 (29) | ||
| Frequently | 375 (35) | ||
| Regularly | 131 (12) | ||
| Hospital | 178 | Respondents per hospital, mean (SD) | 6.0 (6) |
| Hospital size | |||
| <250 beds | 6 (3) | ||
| 250–499 beds | 47 (26) | ||
| 500–999 beds | 73 (41) | ||
| 1000+ beds | 52 (29) | ||
| Country | 31 | Hospitals per country, mean (SD)b | 6 (5) |
Other includes, label (n): social worker (44); ethicist (22); lawyer (21); clergy (20); not otherwise specified (17); patient advocate (7); media (4); psychologist (1)
Countries include: Argentina, Australia, Austria, Belgium, Brazil, Canada, China, Colombia, Cyprus, Czech Republic, Denmark, France, Germany, Hong Kong, India, Ireland, Israel, Italy, Netherlands, Peru, Poland, Portugal, Saudi Arabia, Slovakia, South Africa, Spain, Sweden, Switzerland, Turkey, United Kingdom, and the United States of America
Table 3.
Provider Agreement with Consensus Statements
| Consensus Statement | n (%) |
| Declining to offer CPR when not indicated | |
| Strongly disagree | 15 (1) |
| Disagree | 80 (7) |
| Neutral | 121 (11) |
| Agree | 471 (44) |
| Strongly agree | 381 (36) |
| Permissible to withhold LST in certain circumstances | |
| Strongly disagree | 3 (0) |
| Disagree | 42 (4) |
| Neutral | 55 (5) |
| Agree | 535 (50) |
| Strongly agree | 433 (41) |
| Permissible to withhold LST without consent in certain circumstances | |
| Strongly disagree | 12 (1) |
| Disagree | 85 (8) |
| Neutral | 119 (11) |
| Agree | 598 (56) |
| Strongly agree | 381 (24) |
| Permissible to withdraw LST in certain circumstances | |
| Strongly disagree | 11 (1) |
| Disagree | 68 (6) |
| Neutral | 72 (7) |
| Agree | 540 (51) |
| Strongly agree | 377 (35) |
| Permissible to withdraw LST without consent in certain circumstances | |
| Strongly disagree | 16 (1) |
| Disagree | 92 (9) |
| Neutral | 146 (14) |
| Agree | 573 (54) |
| Strongly agree | 241 (23) |
For each consensus statement, most of the variability in agreement was due to differences among providers (Figure 1). For example, of the total variability in agreement about declining to offer CPR when not indicated, 98% (95% CI 89, 97%) was explained by differences among providers, 0% (95% CI 0,17%) by differences among hospitals, and 2% (95% CI 0, 21%) by differences among countries. After accounting for provider characteristics and hospital size, most of the variability (that is not explained by these characteristics) was still due to differences among providers.
Figure 1.
Partitioning variability in agreement with consensus statements: includes percentage of total variability in agreement with consensus statements due to differences among countries, hospitals, or providers (without fixed effects for provider-and hospital-level characteristics) and standard deviation (SD, 95% Cl) of the variance components. SD of 100 is equivalent to the difference between two categories of agreement (e.g. agree to strongly agree).
Several provider- and hospital-level characteristics were associated with provider views (Table 4). Providers with more involvement in decision-making about withholding or withdrawing LST had higher agreement with each of the 5 consensus statements. Women had lower agreement with the statements than men, although the difference was not statistically significant for all statements. Age and profession were also associated with provider opinion, but both the direction and significance of the association varied by consensus statement. Larger hospital size was associated with less agreement, especially about the permissibility of withholding and withdrawing LST.
Table 4.
| Consensus Statement | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Declining to offer CPR when not indicated | Permissible to withhold LST in certain circumstances | Permissible to withhold LST without consent in certain circumstances | Permissible to withdraw LST in certain circumstances | Permissible to withdraw LST without consent in certain circumstances | ||||||
| Characteristic | Direction of Agreement [Coefficient (95% CI)] | |||||||||
| Age, per 10 years | ↑ | 9 (2, 15) | ↔ | 4 (−2, 9) | ↔ | 1 (−6, 7) | ↔ | 0 (−7, 6) | ↔ | −3 (−10, 3) |
| Female (vs Male) | ↔ | −11 (−24, 2) | ↔ | −6 (−17, 3) | ↓ | −13 (−25, −1) | ↔ | −11 (−22, 0) | ↓ | −13 (−25, −1) |
| Christian (vs non-Christian) | ↔ | −2 (−14, 10) | ↔ | −4 (−14, 5) | ↔ | 1 (−11, 12) | ↔ | 2 (−10, 14) | ↔ | 2 (−10, 14) |
| Profession | poverall < 0.01 | poverall > 0.05 | poverall < 0.05 | poverall < 0.05 | poverall > 0.05 | |||||
| Nurse (vs Physician) | ↓ | −24+ (−38, −9) | ↔ | 2 (−9, 14) | ↔ | 1 (−12, 14) | ↔ | 6 (−8, 19) | ↔ | −3 (−17, 12) |
| Other (vs Physician) | ↔ | −9 (−31, 10) | ↔ | −13 (−30, 3) | ↓ | −22 (−42, −4) | ↓ | −20 (−39, −3) | ↔ | −12 (−32, 6) |
| Other (vs Nurse) | ↔ | 14 (−8, 4) | ↔ | −15 (−33, 2) | ↓ | −24 (−44, −4) | ↓ | −26+ (−46, −7) | ↔ | −9 (−30, 11) |
| Involvement in decisions to withhold/withdraw LST, per higher involvement category | ↑ | 7# (1, 14) | ↑ | 11# (6, 16) | ↑ | 9+ (3, 15) | ↑ | 9# (3, 15) | ↑ | 8+ (2, 14) |
| Hospital size, per larger category | ↔ | −5 (−13, 3) | ↓ | −12# (−18, −4) | ↓ | −10+ (−17, −2) | ↓ | −11 (−18, −1) | ↔ | −8 (−16, 1) |
Model with all fixed effects, intercept, and random effects. Coefficient is difference in mean agreement score associated with unit difference in covariate, when other covariates held constant; presented as coefficient, 95% CI. A positive coefficient means more agreement with the statement; a negative coefficient means less.
Bold: p-value<0.05; Bold+: p-value<0.01; Bold#: p-value<0.001; poverall = p-value for overall test of any difference among categories.
Discussion
Decision-making about end-of-life care involves consideration of the values of patients and their surrogate decision-makers, and because each patient’s situation is unique, some variability is expected. However, when factors operating independently of the patient are major contributors to this variability, concern is warranted. Prior studies have attributed variation in end-of-life practices in the ICU to one or more of three different levels: providers, hospital culture, or national norms (1–12). If hospital culture or regional differences were the primary agents of variability, one would expect end-of-life care in one hospital or within a single country to be more uniform. However, this is often not the case (4, 5, 7, 20). Using data from the WELPICUS study (17), we were able to demonstrate that, although hospital characteristics and country account for some of the variability in agreement about end-of-life care, most of the variance exists at the level of individual providers.
Provider-driven variability generates ethical challenges, if it means that end-of-life care may differ from patient to patient depending on the provider and the particular values they espouse (21). Initiatives to improve end-of-life care in a region or within a specific institution may be most effective when focused on supporting a consensus approach to fundamental elements of end-of-life care. Potential interventions might include development of policies regarding withholding or withdrawing of life-sustaining treatments or the approach to discussing and documenting code status and intensity of care. These policies would need to be grounded in the best available evidence and developed through a consensus process that includes a broad range of clinicians involved in end-of-life care, as well as patient and family stakeholders.
Additional research is needed to understand the underlying reasons for provider-driven variability in end-of-life care. We identified associations between several provider characteristics and agreement with consensus statements about end-of-life care, including: age, sex, profession, and experience with end-of-life care. Other studies have also described differences in practice patterns and found divergent opinions and practice patterns in end-of-life care according to provider age, sex, profession, or experience level (22–26). Experience with decisions to forgo life-sustaining treatment was a predictor of provider agreement for all five consensus statements, which is consistent with existing literature (22, 24–26). It is important to acknowledge that these characteristics are likely only a few of many provider-level factors which play an important role in shaping practice patterns and influencing each individual’s approach to end-of-life care.
The importance of provider-related variability does not diminish the role of organizational factors in influencing treatment intensity at the end of life (6, 27, 28). We found that providers from larger hospitals were less likely to agree with statements affirming the permissibility of withholding or withdrawing LST. Hospital size may correlate with the availability of critical care services and ICU beds, an organizational factor which may influence local attitudes towards aggressive interventions in seriously ill patients (29). In addition, cultural norms within institutions support more or less aggressive approaches to care at the end of life (30), and the importance of recognizing this at the hospital-level is necessary to reduce variability in end-of-life care across centers.
Our study has several important limitations. First, survey results were obtained using a convenience sample, with a relatively small number of respondents per country and hospital, and our results may be subject to response bias. Second, consensus statements developed in WELPICUS may have generated variability beyond that which occurs in normal practice. Participants may have had difficulty interpreting compound statements, given the complexity of the subject matter. Finally, surveys assessed provider agreement with consensus statements about end-of-life care and not actual practice patterns. However, a provider’s perspective on ethical issues plays an important role in shaping the care they provide, even in the absence of a perfect match between perspectives and care.
End-of-life care varies throughout the world. Although some variability is related to hospital culture or regional norms, most is attributable to individual differences between providers. Understanding the primary source of variability in end-of-life care is an important step to help providers engage in efforts to improve care for critically ill patients and their family members. By establishing an approach to key elements of end-of-life care that is more consistent across providers, we may attenuate ethical concerns related to variation in care and also improve patient- and family-centered outcomes.
Supplementary Material
Copyright form disclosure:
Dr. De Robertis received funding from Masimo and Aguettant. Dr. Kross received funding from the National Institutes of Health. Dr. Michalsen received funding from lectures from ViDia Hospital, Stuttgart Hospital, and Konstanz Hospital. Dr. Sprung’s institution received funding from Asahi Kasei Pharma America Corporation (DMC Committee) and LeukoDx (consultant and PI of study). The remaining authors have disclosed that they do not have any potential conflicts of interest.
References
- 1.Cook DJ, Guyatt GH, Jaeschke R, et al. Determinants in Canadian health care workers of the decision to withdraw life support from the critically ill. Canadian Critical Care Trials Group. JAMA : the journal of the American Medical Association 1995;273(9):703–708. [PubMed] [Google Scholar]
- 2.Prendergast TJ, Claessens MT, Luce JM. A national survey of end-of-life care for critically ill patients. American journal of respiratory and critical care medicine 1998;158(4):1163–1167. [DOI] [PubMed] [Google Scholar]
- 3.Azoulay E, Metnitz B, Sprung CL, et al. End-of-life practices in 282 intensive care units: data from the SAPS 3 database. Intensive care medicine 2009;35(4):623–630. [DOI] [PubMed] [Google Scholar]
- 4.DeCato TW, Engelberg RA, Downey L, et al. Hospital variation and temporal trends in palliative and end-of-life care in the ICU. Critical care medicine 2013;41(6):1405–1411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kross EK, Engelberg RA, Downey L, et al. Differences in end-of-life care in the ICU across patients cared for by medicine, surgery, neurology, and neurosurgery physicians. Chest 2014;145(2):313–321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Quill CM, Ratcliffe SJ, Harhay MO, et al. Variation in decisions to forgo life-sustaining therapies in US ICUs. Chest 2014;146(3):573–582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hart JL, Harhay MO, Gabler NB, et al. Variability Among US Intensive Care Units in Managing the Care of Patients Admitted With Preexisting Limits on Life-Sustaining Therapies. JAMA internal medicine 2015;175(6):1019–1026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mark NM, Rayner SG, Lee NJ, et al. Global variability in withholding and withdrawal of life-sustaining treatment in the intensive care unit: a systematic review. Intensive care medicine 2015;41(9):1572–1585. [DOI] [PubMed] [Google Scholar]
- 9.Phua J, Joynt GM, Nishimura M, et al. Withholding and withdrawal of life-sustaining treatments in intensive care units in Asia. JAMA internal medicine 2015;175(3):363–371. [DOI] [PubMed] [Google Scholar]
- 10.Sprung CL, Cohen SL, Sjokvist P, et al. End-of-life practices in European intensive care units: the Ethicus Study. JAMA : the journal of the American Medical Association 2003;290(6):790–797. [DOI] [PubMed] [Google Scholar]
- 11.Dzeng E, Colaianni A, Roland M, et al. Influence of institutional culture and policies on do-not-resuscitate decision making at the end of life. JAMA internal medicine 2015;175(5):812–819. [DOI] [PubMed] [Google Scholar]
- 12.Dzeng E, Dohan D, Curtis JR, et al. Homing in on the Social: System-Level Influences on Overly Aggressive Treatments at the End of Life. Journal of pain and symptom management 2018;55(2):282–289.e281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Truog RD, Campbell ML, Curtis JR, et al. Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College [corrected] of Critical Care Medicine. Critical care medicine 2008;36(3):953–963. [DOI] [PubMed] [Google Scholar]
- 14.Davidson JE, Aslakson RA, Long AC, et al. Guidelines for Family-Centered Care in the Neonatal, Pediatric, and Adult ICU. Critical care medicine 2017;45(1):103–128. [DOI] [PubMed] [Google Scholar]
- 15.Carlet J, Thijs LG, Antonelli M, et al. Challenges in end-of-life care in the ICU. Statement of the 5th International Consensus Conference in Critical Care: Brussels, Belgium, April 2003. Intensive care medicine 2004;30(5):770–784. [DOI] [PubMed] [Google Scholar]
- 16.Lanken PN, Terry PB, Delisser HM, et al. An official American Thoracic Society clinical policy statement: palliative care for patients with respiratory diseases and critical illnesses. American journal of respiratory and critical care medicine 2008;177(8):912–927. [DOI] [PubMed] [Google Scholar]
- 17.Sprung CL, Truog RD, Curtis JR, et al. Seeking worldwide professional consensus on the principles of end-of-life care for the critically ill. The Consensus for Worldwide End-of-Life Practice for Patients in Intensive Care Units (WELPICUS) study. American journal of respiratory and critical care medicine 2014;190(8):855–866. [DOI] [PubMed] [Google Scholar]
- 18.JJ F Extending the linear model with R Generalized linear, mixed effects and noparametric regression models. Boca Raton: CRC Press, 2016. [Google Scholar]
- 19.Goldstein H, Browne W, Rasbash J. Partitioning Variation in Multilevel Models. Understanding Statistics 2002;1(4):223–231. [Google Scholar]
- 20.Wunsch H, Harrison DA, Harvey S, et al. End-of-life decisions: a cohort study of the withdrawal of all active treatment in intensive care units in the United Kingdom. Intensive care medicine 2005;31(6):823–831. [DOI] [PubMed] [Google Scholar]
- 21.Wilkinson DJ, Truog RD. The luck of the draw: physician-related variability in end-of-life decision-making in intensive care. Intensive care medicine 2013;39(6):1128–1132. [DOI] [PubMed] [Google Scholar]
- 22.Christakis NA, Asch DA. Physician characteristics associated with decisions to withdraw life support. American journal of public health 1995;85(3):367–372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hinkka H, Kosunen E, Metsanoja R, et al. Factors affecting physicians’ decisions to forgo life-sustaining treatments in terminal care. Journal of medical ethics 2002;28(2):109–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kelly WF, Eliasson AH, Stocker DJ, et al. Do specialists differ on do-not-resuscitate decisions? Chest 2002;121(3):957–963. [DOI] [PubMed] [Google Scholar]
- 25.Lofmark R, Nilstun T, Cartwright C, et al. Physicians’ experiences with end-of-life decision-making: survey in 6 European countries and Australia. BMC medicine 2008;6:4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Frost DW, Cook DJ, Heyland DK, et al. Patient and healthcare professional factors influencing end-of-life decision-making during critical illness: a systematic review. Critical care medicine 2011;39(5):1174–1189. [DOI] [PubMed] [Google Scholar]
- 27.Lin CY, Farrell MH, Lave JR, et al. Organizational determinants of hospital end-of-life treatment intensity. Medical care 2009;47(5):524–530. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Benoit DD, Jensen HI, Malmgren J, et al. Outcome in patients perceived as receiving excessive care across different ethical climates: a prospective study in 68 intensive care units in Europe and the USA. Intensive care medicine 2018;44(7):1039–1049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Teno JM, Gozalo P, Khandelwal N, et al. Association of Increasing Use of Mechanical Ventilation Among Nursing Home Residents With Advanced Dementia and Intensive Care Unit Beds. JAMA internal medicine 2016;176(12):1809–1816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Dzeng E, Colaianni A, Roland M, et al. INfluence of institutional culture and policies on do-not-resuscitate decision making at the end of life. JAMA internal medicine 2015;175(5):812–819. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.