Introduction

Information about the attitudes of physicians to triage could be used to improve resource allocation in intensive care. The pressure to admit/discharge intensive care unit (ICU) patients varies, depending on the size, location, characteristics and bed occupancy of different hospitals [1]. Reliable and objective scores for triage decisions are unavailable. Cultural differences and religion are also important in medical, ethical decisions [2]. Physicians’ compliance with guidelines for ICU utilization is poor; only 4/20 recommendations of the Society of Critical Care Medicine for ICU triage were observed in a multi-center trial [3]. Physicians’ abilities correctly to assess ICU patient survival at the time of triage is limited [4]. Thus, decisions regarding ICU provision are not based on benefit estimates, allowing admission of patients with little likelihood of survival [2, 5].

The present study evaluated the attitudes of Israeli intensivists to ICU triage and compared them to those of intensivists elsewhere [2, 5] in order to investigate whether differences regarding these issues exist.

Materials and methods

An anonymous survey was conducted among Israeli ICU physicians following IRB approval. Informed consent was implied by questionnaire completion. Questionnaires (with an introductory letter) were mailed to all physician members of the Israel Society of Critical Care Medicine (ISCCM) in June 1994. Three rounds of identical questionnaires were resent to non-responders at 6-month intervals.

The questionnaires were originally developed in English and validated prior to use [5]. They underwent English to Hebrew translation with linguistic validation for the purpose of this study, verifying comprehension and allowing cohort comparison. Data included: (1) the demographic and professional characteristics of the respondents, (2) the importance of various factors in forming a decision whether to admit a patient into the ICU or not and (3) scenarios where the respondents were requested to choose which patients warranted admission to the last ICU bed and which they would discharge from the ICU to accommodate another acutely ill patient. All questions were closed.

Data analysis was performed using SPSS 10.0 (SPSS, Chicago, Illinois). Descriptive statistics (counts and percentages) were used for demographic data and decisions to admit or discharge patients. Chi-square was used to compare the physicians’ responses in relation to their demographic details. Religiosity was used both as an ordered variable and a dichotomy. Spearman Correlation Coefficient (SCC) was chosen to evaluate the association between ordered variables in the physician’s demographic data (experience, seniority, percent of time spent working in ICU, number of hospital and ICU beds) and replies to the various questions regarding admission and discharge of patients. Significance was defined as p less than 0.05 in both Chi-square and SCC.

Confidence intervals were calculated for the difference between the two cohorts in the percent of replies affirming that a factor is important in deciding whether to admit a patient to the last ICU bed. Results were tabulated and presented alongside the previous US study [6].

Results

The response rate was 45% (43/95). The characteristics of the respondents have been published elsewhere [6]. The mean age of the respondents was 50 years (range 35–65). They practised in hospitals with a median total number of 650 beds and 6 ICU beds. Twenty-four respondents (56%) were academic.

Table 1 demonstrates the factors considered important in forming the decision whether to admit a patient to the ICU. Religious physicians placed greater importance on “your personal attitude” than those who were not (p=0.035, Chi-square). Physicians spending more time in the ICU considered the quality of life as viewed by the patient (SCC −0.4062; p<0.05) and the patients’ degree of alertness (SCC −0.4463; p<0.01) less important.

Table 1 Percent of respondents who considered the specific detail important in deciding admission to the last ICU bed
Full size table

Given a situation with empty ICU beds and a patient being admitted to a ward if not granted ICU admission, a similar percent of Israeli (98%) and US (94%) respondents would admit a patient who might live for several years but whose quality of life would be poor according to the physician. An equal percent (77%) of respondents in both cohorts would admit a patient who might live for several years but whose quality of life would be poor according to the patient. More Israeli (70%) than US professionals (54%) were likely to admit a terminally ill patient with a predicted survival of no more than a few weeks. Physicians working in larger (SCC 0.4053; p<0.01) and general (Chi-square p=0.036) ICUs were less likely to admit the terminally ill patient.

Table 2 lists responses regarding admissions to the last ICU bed and discharges in order to accommodate a young asthmatic patient with acute respiratory failure. Israeli intensivists working in larger hospitals (SCC 0.3438; p<0.05) and larger ICUs (SCC 0.348; p<0.05) were less likely to admit a terminal patient with an iatrogenic complication and more likely to discharge a comatose patient (SCC 0.3604; p<0.05 and SCC 0.3577; p<0.05, respectively). Older physicians were less likely to admit a patient suffering from HIV (SCC −0.4463; p<0.01). Both Israeli (84%) and US (87%) respondents did not believe age should be a criterion for refusing ICU admission.

Table 2 Part 1 of the table lists responses to the question “Which patients would you admit to the last ICU bed”. Part 2 of the table lists responses to the question “Which patients would you discharge from the intensive care unit in order to accommodate a 25-year-old asthmatic patient with acute respiratory failure in the setting of a shortage of beds?”. All values given are in percent
Full size table

Discussion

Israeli physicians were less likely than US physicians to admit patients to the ICU. Nonetheless, the majority of respondents would still admit either a patient whose quality of life would be poor according to the patient or a patient with a predicted survival of no more than a few weeks. Almost all physicians would admit a patient whose quality of life would be poor according to the physician.

Factors having an important influence on forming the decision to admit a patient were reminiscent of those described by the US respondents, but agreement was greater in the current study. Similar to the US cohort, most Israeli physicians refused to discharge most ICU patients when presented with a young asthmatic patient with acute respiratory failure that required ICU care in the setting of bed shortage. Notable exceptions were a brain dead child who was a potential organ donor and a patient comatose for 2 weeks.

Comparing the present and US respondents, the majority were male, married and spend a significant amount of time in the ICU [5]. Differences between cohorts included an older age (50 vs 38 years), a physicians-only cohort (the US study included a small number of ICU nurses) and fewer females (5 vs 23%) [5]. Comparison with European respondents was impossible due to questionnaire differences [2].

Patient autonomy was less important for Israelis [5]; their personal attitude and view of the patients’ quality of life was considered as important as that of the patient. European surveys display similar paternalistic attitudes [7, 8]. Most Israeli and European [2] ICU physicians, but only approximately 50% of US respondents [5], were likely to admit a patient with a predicted survival of no more than a few weeks to the last ICU bed. This possibly reflects differences in attitudes towards terminal illness, or diverse expectations from ICU therapy between countries.

Communication barriers exist between ICU physicians and patients [9]. The current study suggests that even when patient preferences are known, physicians may choose not to comply. Few Israeli physicians were likely to discharge a patient with a reversible disease who requested to die, compared with approximately 50% of US professionals. Cultural differences regarding the sanctity of life, a fundamental Jewish principle [10], may have led to reluctance in cooperating with requests inconsistent with the preservation of life. Similar attitudes are more commonly seen in Southern Europe [2] than in Northern Europe [2] and the US [5].

Intensive care unit physicians have a greater obligation to current ICU patients than to admission candidates [11]. Whilst delayed transfer to the ICU increases 30-day mortality [12] and ICU admission decreases hospital mortality [13, 14], early ICU discharge increases mortality [15]. Israelis were both more likely to discharge one patient in order to admit another and less likely to admit new patients than their US counterparts. This may relate to the scarcity of ICU beds in Israel. Additionally, lesser compliance with admission in general may be explained by decreased apprehension with regard to litigation, even in cases of an iatrogenic complication.

Increased Israeli compliance was seen for admission of a patient in lieu of a potential donor. This may relate to the lower incidence of consent for donations in the Israeli public (yielding efforts to maintain a potential donor in the ICU less productive), to Israeli intensivists’ non-acceptance of the organ donor as the sum of his potential rather than as an individual or to greater ICU bed scarcity in Israel. This issue has yet to be examined in Europe.

Study limitations include: the reporting of attitudes rather than actual practice; small sample size and non-representation of all Israeli intensivists. Positive factors include a response rate akin to the US (52%) [5] and Europe (40%) [2], and that a majority of the respondents are ICU directors/senior physicians who are responsible for triage decisions.

Intensive care unit physicians practising outside the US may express alternative attitudes to resource distribution, based on personal ethics or their practice circumstances (e.g. size and type of ICU). This study highlights the existence of such diversity. Further research in individual countries is necessary to improve consistency in medical practice.