Do-not-attempt resuscitation policy reduced in-hospital cardiac arrest rate and the cost of care in a developing country

ABSTRACT

We aim to study the characteristics and outcomes of patients with a Do-Not-Attempt Resuscitation and to determine its impact on the Cost of In-Hospital Cardiac Arrest. A retrospective study of all adult patients admitted to the hospital from June 2021 to May 2022 who had a Do-Not-Resuscitate order. We abstracted patients’ socio-demographics, physiologic parameters, primary diagnosis, and comorbidities from the electronic medical records. We calculated the potential economic cost using the median ICU length of stay for the admitted IHCA patients during the study period. There were 28,866 acute admissions over the study period, and 788 patients had DNR orders. The median (IQR) age was 71 (55-82) years, and 50.3% were males. The most prevalent primary diagnosis was sepsis, 426 (54.3%), and cancer was the most common comorbidity. More than one comorbidities were present in 642 (80%) of the cohort. Of the DNR patients, 492 (62.4%) died, while 296 (37.6%) survived to discharge. Cancer was the primary diagnosis in 65 (22.2%) of those who survived, compared with 154 (31.3%) of those who died (P = 0.002). Over the study period, 153 patients had IHCA and underwent CPR, with an IHCA rate of 5.3 per 1,000 hospital admissions. Without a DNR policy, an additional 492 patients with cardiac arrest would have had CPR, resulting in an IHCA rate of 22.3 per 1000 hospital admissions. Most DNR patients in our setting had sepsis complicated by multiple comorbidities. The DNR policy reduced our IHCA incidence by 76% and prevented unnecessary post-resuscitation ICU care.

KEYWORDS:

• DNR
• IHCA
• cost benefit
• developing country

1. Introduction

In-hospital cardiac arrest (IHCA) is a significant burden on healthcare systems because of the generally poor outcomes and the high costs of ongoing care for those who survive with impaired functional status.

Cardiopulmonary resuscitation (CPR) and post-resuscitation care are the mainstay of management of IHCA. Cardiopulmonary resuscitation is the process that restores cardiac functions and supports ventilation during a cardiopulmonary arrest. It involves the performance of closed-chest compressions, Intubation and ventilation, and electrical defibrillation of the heart in the event of cardiopulmonary arrest. When it was first introduced into medical practice in the 1960s, nearly every dying person was presumed to require CPR. However, with reported survival rates ranging from 4% to 30% for IHCA, It soon became apparent that CPR was prolonging death rather than life [1–4]. Significant side effects were also reported, including considerable resource utilisation and harm to the patients and healthcare providers [5–7]. Developments in bioethics and law made many countries adopt a Do-Not-Resuscitate (DNR) policy. The policy rests on the two pillars of patients’ autonomy and doctors’ rights to determine the futility of treatment. Many countries across the globe now have statutes that support DNR in cases where CPR is likely to be futile, or there is an expressed wish for DNR by the patient or their surrogate [8,9]. The reported rate of DNR in admitted hospital patients varied from 28% to 90%, reflecting cultural norms and institutional practices [10–13]. However, many developing countries, including the UAE, lacked legislation and professional guidelines on DNR, making the practice uncommon in those countries [1,14–16].

Following legislation by the Government of Abu Dhabi, United Arab Emirates, the Department of Health of Abu Dhabi introduced a DNR policy to all government hospitals in April 2020. [Table 1]. The policy stipulates that three consultants (including the responsible consultant) should sign the DNR if CPR is likely to have a poor outcome and will not benefit the patient following a cardiac arrest. The law does not allow the patient or their family to request for DNR in advance, and CPR should still be done if the patient or their family expressly demand it. The DNR was discussed at the time of admission in all critically sick patients and those who were elderly and infirm, thus, in part, aligning with the European Resuscitation Council recommendations on the ethics of resuscitation [17]. Before the policy, and similar to many developing countries, DNR was not practised in the UAE, and all patients who had cardiac arrest were given CPR. An earlier report on In-hospital cardiac arrests in our setting highlighted the impact of a lack of DNR policy on IHCA outcomes [18]. The do-not-resuscitate policy has now been in operation for two years at our institution. We aim to study the characteristics and outcomes of patients with a DNR and assess its impact on IHCA rates and resource utilisation in our setting.

Table 1. Department of health DNR policy.

	Policy purpose
	To guide hospital staff towards practical, safe, and legal application of the provisions of Article 11 of the UAE federal Medical Liability Law # 4 for the year 2016 (Reference 8.1) concerning Don Not Resuscitate (DNR) and therefore not providing cardiac pulmonary resuscitation (CPR) when certain criteria are met.
	Policy statement
	Patients with advanced medical condition(s) such as, but not limited to, advanced cancer, end-stage heart, lung, liver disease or patient with severe injuries incompatible with life shall be allowed to die a natural death without artificial life support in a peaceful and dignified manner in a hospital setting.
	Do not resuscitate (DNR) shall only be processed through this policy to ensure compliance with Article # 11 of the UAE Federal Medical Liability Law #4 for the Year 2016, which states that “Natural death may be allowed by not providing cardiopulmonary resuscitation for dying patient, “when All of the following conditions are met:
	1.	The patient is suffering from an incurable disease.
	2.	If all treatment methods have already been used.
	3.	If it is proved that treatment in this case is futile.
	4.	The attending physician advises not to do cardiopulmonary resuscitation.
	5.	If at least three consultant physicians decide (including the attending physician, provided that they are a consultant) that it is in the best interest of the patient to allow natural death without performing cardiopulmonary resuscitation.
	Procedure
1.	Approval of the patient, patient guardian or custodian is not required when the DNR is decided by three consultant physicians as per above. However, informing the family regarding the DNR decision shall be communicated and documented.
2.	It is prohibited to refrain from doing cardiopulmonary resuscitation in adult patients if explicitly requested by the patient, even if the treatment is futile. The same applies for a paediatric patient when any of the two parents or legal guardians explicitly request cardiorespiratory resuscitation.

2. Patients and methods

2.1. Study design

This is a retrospective study of all adult patients admitted to Tawam Hospital, the United Arab Emirates, from the 1st of June 2021 to the 31st of May 2022 who had a DNR order during the hospital admission.

2.2. Setting

Tawam Hospital is one of the two university teaching hospitals that serve a population of 750,000 in the Al Ain Region of Abu Dhabi Emirate in the UAE [19]. It has 450 in-patient beds, a full complement of tertiary services, and is the regional oncology centre.

2.3. Study population

We included all patients 18 years or older admitted over the study period with a DNR order (ICD-10 code Z66). The study covered the second 12 months following the introduction of the DNR policy to ensure that the new policy was well embedded within our practice.

2.4. Studied variables

Patients’ socio-demographics, physiologic parameters, primary diagnosis, and comorbidities were abstracted from the electronic medical records. The primary outcome measure is the outcome of the patients with DNR orders.

3. Ethical considerations

Tawam Human Research Ethics Committee (T-HREC) approved this study (Ref No: KD/AJ/853)

The cost of intensive care was calculated based on the diagnosis-related group (DRG), a validated case-mix complexity system of healthcare costs [20].

We calculated the intensive care unit (ICU) cost by multiplying the ICU cost per day by the median ICU length of stay (ICU-LOS) for patients with IHCA who achieved return of spontaneous circulation and were admitted to the ICU over the study period, multiplied by the total number of DNR patients who died. We corrected the final cost by 40%, which was the rate of ROSC for IHCA patients in our previous report [18].

ICU cost = ICU cost per day X median ICU-LOS X DNR patients that died X 0.4

4. Statistical analysis

Data are presented as a proportion for categorical variables, median and inter-quarter range (IQR) for continuous variables. A p-value of 0.05 was accepted as statistically significant. Analysis was done using Statistical Package for Social Sciences (IBM, SPSS version 26, Chicago, Il).

5. Results

There were 28,866 acute admissions over the study period. One hundred and fifty-three patients had IHCA and underwent CPR, giving an IHCA incidence of 5.3 per 1,000 hospital admissions. In addition, 788 patients had DNR orders.

Table 2 shows the DNR patients’ demographics and admission physiological parameters by outcomes. The median (IQR) age was 71 (55–82) years, and 50.3% were males. Overall, 296 (37.3%) of the patients survived to discharge. Of the survivors, significantly more patients (71.3%) were admitted to other wards compared with 28.7% who were admitted to the oncology ward (p < 0.001). Only 15% of the patients came from a long-term care facility. The DNR patients had vital signs within the normal range. However, the two groups had significant differences in the median hospital length of stay (LOS) and the GCS. The median (IQR) GCS was 14 (10–15) in those who survived compared with 15 (11–15) in those who died (p < 0.001). The median (IQR) length of stay in the survivors was 11.9 (7.4–21.0) compared with 10.6 (5.1–21.3) in those who died. (p = 0.006).

Table 2. Compares those who survived with those who died.

Variable	Overall	Died N(%) = 492 (100)	Survived N(%) = 296 (100)	P value
Sex (%)				0.941
Male	396 (50.3)	248 (50.4)	148 (50.0)
Female	392 (49.7)	244 (49.6)	148 (50.0)
Age (IQR)	71 (55–82)	68 (53–80)	77 (61–85)	<0.001
Nationality (%)				0.043
Emirati	320 (40.6)	186 (37.8)	134 (45.3)
Non-Emirati	468 (59.4)	306 (62.2)	162 (54.7)
Long-Term care (%)				<0.001
Yes	122 (15.5)	51 (10.4)	71 (24.2)
Admitting Service (%)				<0.001
Oncology	296 (37.6)	211 (42.9)	85 (28.7)
Other Wards	492 (62.4)	281 (57.1)	211 (71.3)
Admission vital signs (IQR)
Pulse	94 (79–110)	94 (80–110)	92 (78–109)	0.140
SBP	113 (100–130)	114 (99–130)	112 (101–128)	0.765
RR	20 (18–24)	20 (18–25)	20 (18–24)	0.852
SpO2	98 (96–100)	98 (96–99)	98 (96–100)	0.500
GCS	15 (10–15)	15 (11–15)	14 (10–15)	<0.001
Length of Stay	11.2 (6.3–21.3)	10.6 (5.1–21.3)	11.9 (7.4–21.5)	0.006

Table 3 shows the primary diagnosis and comorbidity load. The most prevalent primary diagnosis was sepsis, 426 (54.3%), followed by malignancy, 219 (27.9%) and stroke, 38 (4.8%). Respiratory and Cardiac causes were 37 (3.7%) and 11 (1.4%), respectively. Cancer was the primary diagnosis in 65 (22.2%) of those who survived, compared with 154 (31.3%) of those who died (p = 0.002). Overall, 18% of the patients had zero comorbidity, 62.2% had one to three comorbidities, and 19.8% had four or more comorbidities. Figure 1 shows the proportion of those who died increased with increasing comorbidity load, but the difference was not statistically significant.

Figure 1. Shows an increased proportion of dead patients as comorbidity load increases.

Table 3. Primary diagnosis and comorbidity load.

Variable	Overall	Died N(%) = 492 (100)	Survived N(%) = 296 (100)	P value
Primary diagnosis (%)				0.002
Sepsis	426 (54.3)	263 (53.7)	162 (55.3)
Malignant neoplasm	219 (27.9)	154 (31.3)	65 (22.2)
Neurology/Stroke	38 (4.8)	23 (4.7)	15 (5.1)
Respiratory	37 (4.7)	18 (3.7)	19 (6.5)
Cardiac	11 (1.4)	8 (1.6)	3 (1.0)
Renal	4 (0.5)	1 (0.2)	3 (1.0)
Others	54 (6.9)	25 (5.1)	29 (9.8)
Comorbidity Load (%)				0.560
0	141 (18.0)	98 (19.9)	43 (14.7)
1–3	487 (62.2)	295 (60.1)	192(65.8)
>3	155 (19.8)	98 (19.9)	57 (19.5)

Table 4 shows the comorbidities and their frequencies. The most common comorbidities were malignancies (hematologic and solid tumors), cardiac (including ischemic heart disease and heart failure), chronic kidney disease, stroke, and respiratory failure.

Table 4. Comorbidity frequencies.

Comorbidity	N(%)
ONCOLOGY	351 (44.5)
HEART FAILURE	244 (31.0)
CKD	185 (23.5)
STROKE	181 (23.0)
RESPIRATORY FAILURE	167 (21.2)
Others	140 (17.8)
DM	132 (16.8)
Liver disease	50 (6.3)
HIV/TB	5 (0.6)

Over the study period, an additional 153 patients had IHCA and were given full advanced life support resuscitation following the American Heart Association protocol. Among these, 96 (62.7%) patients were admitted to the ICU for post-resuscitation care following ROSC. The median ICU length of stay was two days. The ICU cost per day over the study period in our setting was $3,900. Four-hundred and ninety-two patients with a DNR had a cardiac arrest but did not have CPR over the study period. Had they been resuscitated, it would have resulted in an IHCA rate of 22.3 per 1,000 annual hospital admissions. Assuming a 40% ROSC rate following resuscitation [18], an additional 197 patients (40% of 492) would have required post-cardiac arrest care in the ICU. The extra cost of ICU care for these patients would have amounted to 1,536,600 US dollars (3,900 X 2 X 197).

6. Discussion

This study shows that the implementation of the DNR policy reduced the IHCA rate from 22.3 to 5.3 per 1,000 hospital admissions, representing a drop of 76%. This is similar to IHCA rates reported from Europe and North America but lower than those from Africa and Latin America [1,3,21,22]. The policy also prevented 394 unnecessary ICU admission days with an estimated reduction of more than $ 1.5 million US dollars in the cost of health care expenditure.

The development and evolution of DNR date back to the mid-1970s and are rooted in bioethics, case laws and statutes [23]. The state of New York was the first in the United States to put DNR into law in 1988 before it became widespread throughout the United States [24]. The statutes gave legal backing to the two pillars of DNR: (1) A patient with capacity has the right to refuse treatment, including CPR. (2) Doctors can determine the futility of care without liability. Placing a patient on a DNR does not equate to denying other life-sustaining treatments such as oxygen, intravenous fluids and antibiotics [25,26]. Instead, it is an integral part of the advance care plan, which includes a living will, surrogate decision-making, and treatment escalation policy [27,28]. In this study, the DNR patients had a ceiling of care limited to the ward level and were not admitted to the ICU. The DNR practice in our setting is still developing, and unlike in North America and Europe, individuals are not legally allowed to give an advance directive for DNR. In addition, the patient or their surrogate legally and culturally cannot opt out of resuscitation, but following discussion, they can demand CPR and the clinician will oblige them, though the effort may be futile. Cardiopulmonary resuscitation has significant side effects, such as harm to providers, a burden on caregivers, and the risk of achieving an outcome that is incompatible with the patient’s expressed wish and sociocultural preference. Despite the side effects of CPR and the benefits of the DNR policy, including dying with dignity, creating opportunities for shared decisions, and helping the family with the grieving process [29], many developing countries still do not have DNR statutes [1,14–16,30].

Patients with DNR have reduced hospital length of stay, increased mortality and reduced ICU costs [14,28,31]. This study shows that an additional 493 patients would have had CPR without a DNR policy. Assuming a 40% ROSC rate reported previously from our setting [18], an additional 197 patients would have required post-cardiac arrest ICU care. Using our cost modelling, the savings in the cost of ICU care alone would have been over 1.5 million US dollars. This is similar to reports by others [14,29]. The calculated savings are modest, and the median length of stay of 2 days is likely underestimated for the DNAR group. Low- and middle-income countries need to adopt DNR practices through legislation and a change in clinical practice to ensure that critical resources are used for the patients that are likely to benefit maximally from them.

Sepsis was the most common primary diagnosis in this study. The multiple comorbidities in these patients may have made them prone to sepsis due to immunosuppression. Wang et al. reported that DNR was an independent predictor of mortality in patients with sepsis [32]. The poor outcomes in sepsis patients with DNR may reflect the underlying comorbidities and the ceiling of care, which often preclude critical care. Malignant neoplasms were the second most common primary diagnosis. The high prevalence of malignancies in this study reflects the fact that our institution is the Regional Cancer Centre.

The DNR patients in this study were elderly, and over 40% of them were UAE Nationals. This is an overrepresentation, considering that the UAE Nationals constitute only 10% of the general population [19]. The UAE population is made up of predominantly young migrant workers and expatriates who tend to retire to their home countries. The elderly infirm also tend to prefer Government or family-funded home care to Long-term institutional Care.

7. Limitations

This study has some limitations. Our method of selecting suitable patients for DNR was subjective, based on the consensus of three attending physicians, but an objective approach, such as that proposed in the Good Outcomes Following Attempted Resuscitation (GO-FAR), may have been more helpful for patients and their surrogates when discussing end-of-life care [33]. In addition, we did not report on the outcomes, including the functional status of those patients who were discharged from the hospital alive. Finally, the outcomes in IHCA patients following the introduction of DNR policy in our setting are unknown, but this can be the focus of future studies. However, the 63% ROSC rate in those who were resuscitated was significantly higher than the 40% we previously reported [18], indicating that the policy may have improved our IHCA survival rate.

8. Conclusions

Most DNR patients in our setting had sepsis, complicating multiple comorbidities. The DNR policy reduced our IHCA rate by 76%, preventing unnecessary ICU care in 197 patients. By establishing statutes for DNR, developing countries may improve patients’ experience through advance directives, advanced care plans, and shared decision-making, ensuring critical resources are used for patients most likely to benefit from them. Further research on DNR laws and statutes in developing countries would assist in understanding the current state and pave the way for improvements in end-of-life care.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the no funding [NA].