Article Text

Download PDFPDF
Interventions in hospitalised patients with cancer: the importance of impending death awareness
  1. Eric Cornelis Theodorus Geijteman1,2,
  2. Marcella van der Graaf1,
  3. Frederika E Witkamp1,3,
  4. Sanne van Norden1,
  5. Bruno H Stricker4,
  6. Carin C D van der Rijt1,5,
  7. Agnes van der Heide2 and
  8. Lia van Zuylen1
  1. 1 Department of Medical Oncology, Erasmus MC Cancer Institute, Rotterdam, The Netherlands
  2. 2 Department of Public Health, Erasmus University Medical Center, Rotterdam, The Netherlands
  3. 3 Research Centre Innovations in Care, Rotterdam University of Applied Sciences, Rotterdam, The Netherlands
  4. 4 Department of Epidemiology, Erasmus University Medical Center, Rotterdam, The Netherlands
  5. 5 Netherlands Comprehensive Cancer Organization, Utrecht, The Netherlands
  1. Correspondence to Eric Cornelis Theodorus Geijteman, Department of Medical Oncology, Erasmus MC Cancer Institute, Rotterdam 3000 CA, The Netherlands; e.geijteman{at}erasmusmc.nl

Abstract

Objectives Burdensome and futile interventions with the aim of prolonging life should be avoided in dying patients. However, current clinical practice has hardly been investigated.

We examined the number and type of diagnostic and therapeutic medical interventions in hospitalised patients with cancer in their last days of life. In addition, we investigated if physician awareness of impending death affected the use of these interventions.

Methods Questionnaire study and medical record study. Attending physicians of patients who died in a university hospital between January 2010 and June 2012 were asked whether they had been aware of the patient’s impending death. The use of diagnostic and therapeutic interventions and medications was assessed by studying patients’ charts. We included 131 patients.

Results In the last 72 and 24 hours of life, 59% and 24% of the patients received one or more diagnostic interventions, respectively. Therapeutic interventions were provided to 47% and 31%. In the last 24 hours of life, patients received on average 5.8 types of medication.

Awareness of a patient’s impending death was associated with a significant lower use of diagnostic interventions (48% vs 69% in the last 72 hours; 11% vs 37% in the last 24 hours) and several medications that potentially prolong life (eg, antibiotics and cardiovascular medication).

Conclusions Many patients with cancer who die in hospital receive diagnostic and therapeutic interventions in the last days of life of which their advantages are questionable. To improve end-of-life care, medical care should be adapted.

  • clinical decisions
  • end of life care
  • hospital care
  • prognosis
  • terminal care

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Background

Hospital care is typically focused on cure and prolongation of life. As a result, dying patients may receive interventions that are not aimed at promoting their comfort.1 This has important negative consequences for both patients and their relatives and is therefore in conflict with what has been found to constitute ‘a good death’.2

In almost all patients with advanced incurable cancer, death is preceded by a more or less clear period of imminent dying.3 When it is recognised and acknowledged that the dying phase has started, interventions can and should be focused at providing patients with all the comfort they need and wish.1 4

The purpose of this study was to investigate (1) how many and which diagnostic and therapeutic interventions are applied by physicians in hospital in the last 72 hours of life of inpatients with cancer and (2) whether awareness of impending death of the attending physician is associated with the application of these interventions.

Methods

This study is part of a study to explore and understand palliative and terminal care in the hospital (PalTec-H).5 We studied medical care for inpatients with cancer who died between January 2010 and June 2012 during their stay in Erasmus MC, a 1300-bed university hospital in Rotterdam, the Netherlands. All inpatient wards of the hospital participated, with the exception of the department of psychiatry and the intensive care departments. Attending physicians were asked to fill out a questionnaire within 1 week after a patient had died. We studied the patients’ medical charts. Patients who died within 72 hours of their hospital admission were excluded.

The questionnaire included a question about the physicians’ awareness of a patient’s impending death. Physicians were asked: ‘had it prior to death been clear that the patient would die within hours or days?’ They could answer ‘yes’, ‘more or less’, or ‘no’. We merged the answer options ‘more or less’ and ‘no’ into ‘no’. When physicians answered ‘yes’ they were asked: ‘when did this become clear?’ We defined a physician as having been aware of a patients’ impending death when he answered that it had been clear that the patient would die within hours or days more than 24 hours before the patient actually died. The questionnaire also included a question on whether the patient was treated with a curative or palliative intent.

Two physician researchers (ECTG, MvdG) and one medical student (SvN) reviewed the medical chart of each patient about whom physicians filled out a questionnaire. All diagnostic and therapeutic medical interventions in the last 3 days were registered using a predefined list of interventions, as well as medication use during this period. We assessed whether or not an intervention or medication was used in either of the two periods: the last 72 hours and the last 24 hours of life.

The statistical significance of bivariate associations between physicians’ awareness of a patient’s impending death and the use of diagnostic and therapeutic interventions was analysed with χ2 tests. An alpha of 0.05 was used as the cut-off for significance.

Results

Physicians completed the questionnaire for 150 patients with cancer who died during the study period (physicians’ response rate of the PalTec-H questionnaire study was 45%). Data of 19 patients could not be analysed, either because of missing information on the physician’s awareness of the patient’s impending death or because of an incomplete medical chart. Data about 131 patients could be used for the analysis. In 63 patients (48%), physicians had been aware of the patient’s impending death (web only Supplementary table 1). The mean duration of the final hospital stay was 13 days for both groups (SD 11.0 and 10.3, respectively). The most common cancer types were lung cancer (n=21), cancer of the urological tract (n=17) and haematological cancer (n=15). Ninety-two per cent of all patients were treated with a palliative intent during their stay in the hospital.

Supplementary file 1

Diagnostic interventions

In the last 72 and 24 hours of life, 59% and 24% of the patients, respectively, received one or more diagnostic interventions, mostly blood sampling and radiological procedures (web only Supplementary table 2). Among patients for whom the physician had been aware of their impending death, 48% received one or more diagnostic interventions in the last 72 hours compared with 69% of the patients in the other group (P=0.013). The percentage of patients who received one or more diagnostic interventions in the last 24 hours of life was also significantly lower when the imminence of death had been acknowledged (11% vs 37% (P<0.001)).

Supplementary file 2

Therapeutic interventions

In the last 72 hours and 24 hours of life, 47% and 31% of the patients received one or more therapeutic interventions, respectively (Table 1). Awareness of impending death was not significantly associated with receiving therapeutic interventions in the last 72 and 24 hours, with the exception of intravenous fluids which were used less often in the last 24 hours of life when the physician had been aware of impending death (8% vs 28% (P=0.003)). The interventions most often applied were intravenous fluids and enteral tube feeding. In almost all patients who received enteral tube feeding, this feeding was continued until the patient passed away.

Table 1

Therapeutic interventions in the last days of life

Medication

On average, patients used 7.2 types of medication (SD 4.38) in the last 72 hours of life (web only Supplementary table 3) and 5.8 (SD 4.24) in the last 24 hours of life (web only Supplementary table 4). Patients for whom the physician had been aware of their impending death used fewer medications in the last 24 hours of life than patients for whom the physician had not been aware of their impending death (mean 5.2 vs 6.4, P=0.038). The percentage of patients who used cardiovascular medications and medications for obstructive airway diseases in the last 72 hours and 24 hours of life was significantly lower when the imminence of death had been acknowledged. In the last 24 hours of life, antibiotics were less often prescribed to patients in whom the physician had been aware of impending death. There were no significant differences in the use of other medications.

Supplementary file 3

Supplementary file 4

Discussion

Our study shows that many patients with cancer who died in hospital received diagnostic and therapeutic interventions in the last days of life. Our study also indicates the importance of physicians’ awareness of impending death, as it may reduce the use of—often burdensome and futile—interventions.

Ideally, when the attending physician is aware of the imminence of a patient’s death, burdensome interventions with the aim of prolonging life are avoided.1 6 Although some patients in whom the attending physician had been aware of the imminence of death received one or more diagnostic or therapeutic interventions with the potential of prolonging life, our study shows that acknowledgement of the imminence of death is associated with less diagnostic interventions. This finding is in line with earlier research on end-of-life care in hospitals and other settings of Veerbeek et al in which diagnostic interventions were also found to be applied significantly less often in patients for whom the dying phase was recognised.7

We did not find an association between awareness of patient’s impending death and therapeutic interventions. Veerbeek et al’s findings were similar.7 An important reason for this non-association in our study was that enteral tube feeding was continued in many cases where patient’s imminent death was acknowledged. The use of enteral tube feeding in the dying phase is highly questionable as it may cause harm as a result of complications, such as aspiration and sepsis.8 A reason for continuing enteral tube feeding until the last 24 hours of life might be that its withdrawal may lead to distress for the patient and their family members.9 10

Although extensive diagnostic and therapeutic interventions, such as endoscopy, mechanical ventilation and resuscitation, were used in only a small percentage, many patients received other less extensive interventions in their last days of life, also if their imminent death was acknowledged. There seems to be room for improvement as even less extensive diagnostic and therapeutic interventions could have great impact on dying patients and their relatives.11 A plausible reason for the use of these interventions until shortly before death may be that caregivers just persist in their daily routine of work.7 12

Our study highlights that patients use many medications in the last 72 and 24 hours of life, either with the potential of prolonging life or of symptom control. Awareness of a patient’s impending death slightly decreased the number of medications in the last 24 hours of life. However, both patients in whom the attending physician had or had not been aware of the imminence of death used so-called preventative medications, such as statins, which can be considered inappropriate in the last phase of life as they have no short-term benefit.13 In two recent retrospective studies, it was also shown that many inappropriate medications are continued until shortly before death.14 15 We did not find significant differences in prescriptions of medication to relieve symptoms such as opioids.

Some limitations of this study need to be considered. First, this was a retrospective study which has inherent limitations such as recall bias among physicians when answering the question on whether or not they were aware of the patient’s impending death. Second, selection bias cannot be ruled out as physicians’ response rate of the PalTec-H questionnaire study was 45%. Third, this study was performed in one hospital which limits its generalisability.

Conclusion

Physician’s awareness of hospitalised patients’ impending death is associated with reduced use of diagnostic interventions and life-prolonging medication. However, even patients in whom the physician is aware of their impending death may receive interventions that have doubtful benefits in the last days of life. Further adaptations of medical care in hospital for patients who are in the last days of life are needed.

References

Footnotes

  • Contributors Conception or design of the work: ECTG, MvdG, FEW, CCDvdR, AvdH and LvZ. Data collection: ECTG, MvdG, FEW and SvN. Data analysis and interpretation: ECTG, BHS and AvdH. Drafting the article: ECTG, CCDvdR, AvdH and LvZ. All authors contributed to the critical revision of the article and final approval of the version to be published.

  • Funding Data collection was funded by the Erasmus MC Medical Research Committee.

  • Competing interests None declared.

  • Ethics approval Medical ethical research committee of the Erasmus University Medical Center.

  • Provenance and peer review Not commissioned; externally peer reviewed.