Outcome and critical care resources utilised by do not attempt cardiopulmonary resuscitation (DNACPR) patients admitted to the ICU at a tertiary hospital in Saudi Arabia: a retrospective review of the critical care database | Critical Care

The characteristics of the ICU patients (n = 7,104) are summarized in Table 1. During the study period, 7,104 patients were admitted to the intensive care unit (ICU). Among them, 988 (13.9%) patients had a documented Do Not Attempt Cardio Pulmonary Resuscitation (DNACPR) order after ICU admission, while 6,116 (86.1%) patients were classified as non-DNACPR (full code).

The DNACPR patients were significantly older than the non- DNACPR patients, with a mean age of 61.4 ± 16.2 years versus 55.9 ± 16.6 years, respectively (p < 0.001). Approximately 31% (n = 306) of DNACPR patients and 27% (n = 1,657) of non-DNACPR patients were smokers. DNACPR patients were admitted more frequently from inpatient departments than non- DNACPR patients (52.4% vs. 30.4%, p < 0.001). At admission, the APACHE IV and SOFA scores were significantly higher among DNACPR patients (57.4 ± 34.5 vs. 39.1 ± 26.4, p < 0.001 and 7.5 ± 4.9 vs. 4.4 ± 4.2, p < 0.001, respectively). Several underlying comorbidities were significantly more prevalent among DNACPR patients than non- DNACPR patients, including diabetes (49.7% vs. 45.5%, p = 0.014), cardiac diseases (35.5% vs. 30.5%, p = 0.002), renal failure (21.2% vs. 16.5%, p < 0.001), chronic liver disease (7.0% vs. 2.5%, p < 0.001), and malignancy (42.4% vs. 34.9%, p < 0.001). Additionally, DNACPR patients were significantly more likely to be bedridden compared to non-DNACPR patients (17.6% vs. 11.3%, p < 0.001). The most common reasons for ICU admission among DNACPR patients were sepsis (66.6% vs. 34%, p < 0.001), respiratory failure (52.7% vs. 29.6%, p < 0.001), and hemodynamic instability (44.9% vs. 13.6%, p < 0.001). In contrast, the most common reason for ICU admission among non- DNACPR patients was postoperative care following major elective surgery (34.4% vs. 7.9%, p < 0.001).

DNACPR group results

Concerning the DNACPR patients, the time from ICU admission to DNACPR order was (mean 11.9 ± 17.6), while the time from DNACPR order to death (mean 9.2 ± 16.2), and the number of DNACPR patients died within 24 h from DNACPR order were 210 (21%) from total DNACPR patients. The reasons of DNACPR has been identified among 620 (63%) of DNACPR patients (Table 1). The commonest reasons of DNACPR as following: terminal Illness advanced cancer in 247 (39.8%) patients, multiple sever co-morbidities with poor quality of life (bed ridden) in 148 (23.8%), severe neurological impairment with poor prognosis for recovery in 110 (17.7%), and refractory shock multi-organ failure in 78 (12.6%) (Table 2). All patients signed as DNACPR were on full support of care with exception of minor number of case whom had new intervention after DNACPR order. After DNACPR order; new mechanical ventilation were given to among 28 patients and new CRRT started on 14 patients only.

ICU resource comparing DNACPR vs. non- DNACPR results

DNACPR patients utilized resources significantly more than non- DNACPR patients (Table 3). CRRT utilization was significantly higher among DNACPR patients (283 [28.6%] versus 412 [6.7%]; P < 0.001). Mechanical ventilation utilization was significantly higher among DNACPR patients (878 [88.9%] versus 2530 [41.4%]; P < 0.001). The endoscopy procedure was performed for 55 (5.6%) DNACPR patients and 115 (1.9%) DNACPR patients (P < 0.001). Sedation was administered to 843 (85.3%) DNACPR patients and 2141 (35.0%) non- DNACPR patients; P < 0.001.

With respect to imaging studies and procedures, approximately 590 (59.7%) DNACPR patients underwent CT imaging, whereas 3268 (53.4%) non-DNACPR patients underwent CT imaging (P < 0.001). The median number of CT scans was 3 (2–4 IQR) for DNACPR patients and 2 (1–3 IQR) for non- DNACPR patients. On the other hand, 280 (28.3%) DNACPR patients underwent MRI, and 1203 (19.7%) non- DNACPR patients underwent MRI (P < 0.001). Ultrasound imaging was performed for 298 (30.2%) DNACPR patients and 972 (15.9%) non-DNACPR patients (P < 0.001). The number of interventional radiology procedures was significantly higher among DNACPR patients (206 (20.9%) versus 871 (14.2%), P < 0.001). In addition, echocardiography imaging (ECHO) was significantly more common for DNACPR patients (212 (21.5%) versus 641 (10.5%), P < 0.001).

Regarding antimicrobial therapy utilization, DNACPR patients tended to receive significantly more antimicrobial combinations, with 76.1% of DNACPR patients receiving two or more antibiotics, compared to 56.8% of non-DNACPR patients.

In addition, the utilization of blood and blood products was significantly higher among DNACPR patients. Approximately 149 (15.1%) DNACPR patients received 352 units of PRBCs, whereas 253 (4.1%) non-DNACPR patients received 551 units (P < 0.001). Eighty-three (8.4%) DNACPR patients received 390 units of FFP, whereas 114 (1.9%) non-DNACPR patients received 526 units of FFP (P < 0.001). Approximately 72 (7.3%) DNACPR patients received 640 units of platelets, whereas 82 (1.3%) non-DNACPR patients received 758 units of platelets (P < 0.001). The average length of stay for DNACPR patients was significantly higher compared to non-DNACPR patients (20.4 ± 25.1 versus 8.0 ± 11.2 (P < 0.001)). Additionally, the mortality rate was significantly higher (76.7%) among DNACPR patients compared to non-DNACPR patients (7.7%) (P < 0.001).

Multivariate logistic regression

Table 4 presents multivariate logistic regression analysis after adjustment between the groups for severity of illness after controlling for APACHI IV score and cancer as a comorbidity. The analysis showed that DNACPR patients are significantly receiving, CRRT (AOR 4.4, 95% CI 3.6–5.4), mechanical ventilation (AOR 7.8, 95% CI 6.1–9.9), sedations (AOR 7.5, 95% CI 6.0-9.3), MRI (AOR 1.3, 95% CI 1.1–1.5), ultrasound (AOR 1.7, 95% CI 1.4–2.1), intervention radiology (AOR 1.4, 95% CI 1.6–1.7), ECHO (AOR 2.2, 95% CI 1.8–2.8), endoscopy (AOR 2.1, 95% CI 1.4–3.3), and blood transfusion (AOR 2.9, 95% CI 2.2–3.8) with significant P > 0.001 respectively. However, the utilization of CT and antibiotics was not significantly different between the two groups after adjustment for the differences in severity of illness and cancer as a comorbidity.

Sub-group analysis of emergent admissions (excluding elective planned admissions due to postoperative surgery) comparing DNACPR with non-DNACPR patients (Tables 5, 6, 7, 8 and 9)

A total of 2184 planned admissions were excluded from the analysis to study the emergency admissions subgroup, 2106 patients among the non-DNACPR, while 78 patients among the DNACPR were excluded (Table 5). Comparing DNACPR with non-DNACPR patients’ emergency admissions. DNACPR patients’ emergency admissions were significantly older (mean ± SD: 61.47 ± 16.2 years) compared to non- Comparing DNACPR with non-DNACPR patients (57.10 ± 16.8 years) (P < 0.001). The DNACPR patients had a higher prevalence of Chronic liver disease (7.5% vs. 2.9%, P < 0.001), Cancer (41.5% vs. 23.2%, P < 0.001), and Bedridden status (10.1% vs. 6.5%, P < 0.001). Conversely, hypertension was more prevalent among non-DNACPR (56.3% vs. 60.4%, P = 0.020), while all other comorbidities were comparable between the 2 groups. The most common reasons for admission in both groups were septic shock and respiratory failure; however, DNACPR patients had a significantly higher rate of admissions with septic shock and respiratory failure compared to non-DNACPR patients (68.9% vs. 47.2%, P < 0.001, and 56.6% vs. 43.7%, P < 0.001, respectively). In contrast, non-DNACPR had significantly higher admissions with neurological and cardiovascular emergencies (31.1% vs. 34.9%, P = 0.031, 4.1% vs. 9.9%, P < 0.001, respectively). Post-CPR (10.4% vs. 2.3%, P < 0.001). Post RRT (12.3% vs. 7.8%, P < 0.001). The calculated APACHE IV and SOFA scores upon ICU admission were significantly higher among the DNACPR group (mean ± SD: 68.96 ± 30.53 vs. 46.94 ± 26.518, P < 0.001, and mean ± SD: 7.48 ± 4.8 vs. 4.68 ± 3.84, P < 0.001, respectively). Indicating from the time of admission to ICU, DNACPR patients were more severely ill and with more organ failure than the non-DNACPR group, and a poor outcome is expected from the time of admission. Although all the interventions measured among DNACPR group were collected after DNACPR order date, and despite the multivariate regression adjustment for the differences between groups with regards to severity of illness (APACHI IV), organ dysfunction (SOFA) and cancer comorbidity. the DNACPR group showed a significantly higher utilization of various critical care interventions compared to non- DNACPR group (Table 9), including Mechanical ventilation days: AOR = 4.9 (95% CI: 4.77–5.06, P = 0.0001), CRRT days: AOR = 2.46 (95% CI: 2.04–2.96, P = 0.0001), Ultrasound: AOR = 1.20 (95% CI: 1.03–1.35, P = 0.021), CT: AOR = 1.13 (95% CI: 1.06–1.21, P = 0.001), MRI: AOR = 1.18 (95% CI: 1.03–1.35, P = 0.014), Intervention radiology: AOR = 1.27 (95% CI: 1.09–1.50, P = 0.003), ECHO: AOR = 1.98 (95% CI: 1.58–2.49, P = 0.0001), and Blood Transfusions (PRBCs, Platelets, FFP): All showed significantly higher odds for the number of units including: PRBCs: AOR = 2.35 (95% CI: 1.94–2.85, P = 0.0001), Platelets: AOR = 1.7 (95% CI: 1.47–1.98, P = 0.0001), Fresh frozen transfusion: AOR = 2.31 (95% CI: 1.90–2.80, P = 0.0001). with regards to length of stay and mortality outcomes (Table 7). The mean ICU LOS for DNACPR patients was 20.5 ± 25.3 days, significantly longer than for non-DNACPR patients (9.26 ± 11.9 days) (P < 0.001). the DNACPR patients had a substantially higher ICU mortality rate (77.3% vs. 10.8%, P < 0.001).