Circadian rhythm of hospital deaths: comparison between intensive care unit and non-intensive care unit

Abstracts

OBJETIVO: A demonstração de que a mortalidade cardiovascular segue um ritmo circadiano fez com que fosse verificado se pacientes que falecem em unidade de tratamento intensivo (UTI) e unidade clínica (UC) obedecem a este ritmo. MÉTODOS: Foram analisados todos os óbitos hospitalares ocorridos entre 1º de janeiro de 2006 e 31 de julho de 2010. O padrão circadiano dos óbitos foi analisado dividindo-se as 24 horas do dia em períodos de 2 horas. Utilizou-se o teste do qui-quadrado para comparação de variáveis categóricas e o teste t de Student ou a análise da variância fator único para comparação de variáveis contínuas. Um valor de p < 0,05 foi considerado estatisticamente significante. RESULTADOS: No período estudado, foram registrados 700 óbitos, 211 (30,1%) na UTI e 88 (12,6%) na UC. Houve mais óbitos nas primeiras horas do dia, entre 6 e 12 h, na UC, em comparação à UTI (38% vs. 21%; p = 0,004). Na UTI, 21% dos óbitos ocorreram entre 6 e 12 h, 30% entre 12 e 18 h, 26% entre 18 e 24 h e 24% entre 24 e 6 h (p = 0,13). Enquanto, na UC, 38% ocorreram entre 6 e 12 h, 18% entre 12 e 18h, 19% entre 18 e 24 h e 25% entre 24 e 6 h (p = 0,005). CONCLUSÃO: Na UC, os óbitos ocorrem com maior frequência no período da manhã e seguem um padrão circadiano, enquanto na UTI esse fenômeno não ocorre.

Mortalidade hospitalar; ritmo circadiano; transtornos cronobiológicos

ORIGINAL ARTICLE

Circadian rhythm of hospital deaths: comparison between intensive care unit and non-intensive care unit

Antônio Sérgio Cordeiro da Rocha^I; Mônica Peres de Araújo^II; Augusta Campos^III; Rubens Costa Filho^IV; Evandro Tinoco Mesquita^V; Marcus Vinicius Santos^VI

^IPhD in Cardiology; Universidade de São Paulo (USP); Coordinator, Clinical Unit, Hospital Pró-Cardíaco, Rio de Janeiro, RJ, Brazil

^IIMSc in Cardiology, Universidade Federal Fluminense (UFF); Physician, Hospital Pró-Cardíaco, Rio de Janeiro, RJ, Brazil

^IIIBoard Certified Cardiologist, Sociedade Brasileira de Cardiologia (SBC); Physician, Hospital Pró-Cardíaco, Rio de Janeiro, RJ, Brazil

^IVCoordinator, Intensive Care Unit, Hospital Pró-Cardíaco, Rio de Janeiro, RJ, Brazil

^VPhD in Cardiology, USP; Medical Director, Hospital Pró-Cardíaco, Rio de Janeiro, RJ, Brazil

^VIGeneral Director, Hospital Pró-Cardíaco, Rio de Janeiro, RJ, Brazil

Correspondence to

SUMMARY

OBJECTIVE: The demonstration that cardiovascular mortality follows a circadian rhythm led us to verify whether patients dying at the intensive unit care (ICU) and at the non-intensive unit care (non-ICU) follow that rhythm.

METHODS: All hospital's deaths occurring between January 1, 2006 and July 31, 2010 were analyzed. The circadian pattern of the time of death was analyzed in twelve 2-hour intervals. The Chi-square test was used to compare proportions, and Student's t test or ANOVA single factor to compare continuous variables. A p-value < 0.05 was considered statistically significant.

RESULTS: During the study period 700 deaths occurred in the hospital, 211 (30.1%) at the ICU and 88 (12.6%) at the non-ICU. There were more deaths in the first hours of the day, between 6 am and 12 am, at the non-ICU in comparison to the ICU (38% vs. 21%; p = 0.004). In the ICU, we observed that 21% of the deaths occurred between 6 am and 12 pm, 30% between 12 pm and 6 pm, 26% between 6 pm and 12 am and 24% between 12 am and 6 am (p = 0.13) , whereas, at the non-ICU, 38% of the deaths occurred between 6 am and 12 pm, 18% between 12 pm and 6 pm, 19% between 6 pm and 12 am and 25% between 12 am and 6 am (p = 0.005).

CONCLUSION: At the non-ICU, deaths occur more often in the morning period and follow a circadian rhythm, which does not occur at the ICU.

Keywords: Hospital mortality; circadian rhythm; chronobiological disorders.

Introduction

Because of the Earth's rotation, all living species on the planet's surface are exposed to the standard of 24 hours of light and darkness. In response to this regular variation of the light/dark cycle, living beings have developed an endogenous circadian rhythm that repeats itself every 24 hours. With the knowledge obtained over time, we know that this rhythm is important in maintaining the life of living organisms in response to environmental changes. However, changes in human behavior over the centuries, promoted by the modernization of customs and habits, brought a threat of disruption to this rhythm1-5.

In fact, when observing that cardiovascular mortality follows a rhythm, which accompanies neuroendocrine and metabolic alterations, one understands the importance of this biological process on human nature^6,7. Within this context, several authors have studied the differences in outcomes of patients admitted during the day compared to those admitted during shifts in intensive care units (ICU), but no studies have been dedicated to compare the time of occurrence of deaths in the ICU and the non-intensive care unit (non-ICU)^8-12.

The objective of this study is to observe the differences in time of deaths in an ICU compared with the non-ICU and whether there is a circadian pattern of these deaths occurring in these two hospital units.

Methods

We collected the data on all deaths that occurred in the intensive care unit (ICU) and the non-intensive care unit (non-ICU). Thus, this analysis excluded deaths that occurred in the coronary unit, post-operative adult and child units and the emergency department of the institution.

The following data were collected: demographic characteristics, diagnosis of underlying diseases, direct cause of death and time of occurrence of deaths.

We considered cardiovascular disease as the presence of any one of the following pathologies: coronary artery disease, cerebrovascular disease, peripheral vascular disease or thoracic or abdominal aortic disease documented by clinical, laboratory or angiographic pictures compatible with the diagnosis. As infectious diseases, we considered the presence of clinical and laboratory pictures consistent with the presence of any infectious agent of any etiology and affecting any organ system. As cancer, we considered the presence of documented malignancy affecting any organ or system. Neoplasias were considered as direct causes of deaths whenever there was cachexia or multiple organ metastases. As for pulmonary embolism (PE), angiographic evidence of thrombus involving the pulmonary artery circulation was considered. As dementia, we considered the association of memory impairment and some type of cognitive impairment. As kidney failure, we considered the presence of serum creatinine > 2 mg/dL or need for dialysis. Acute myocardial infarction (AMI) was considered when it met the ACC/ESC criteria¹³. Cerebrovascular accident (CVA) was considered in the presence of clinical evidence associated with imaging consistent with the diagnosis. We also evaluated the percentage of patients who were receiving comfort measures in the two units.

The circadian pattern of occurrence of deaths was obtained by dividing the 24 hours of the day in 2-hour periods.

The data are shown as absolute numbers or proportions. For comparison of continuous variables Student's t test or one-way analysis of variance (ANOVA) was used and the Chi-square test or Fisher's exact test was used to compare categorical variables. The significance level was set at 5% and all calculations were two-tailed.

Results

Between January 1, 2006 and July 31, 2010 there were 700 deaths in the entire institution, corresponding to a hospital mortality rate of 3.6%. Of the 700 deaths, 211 (30.1%) occurred in the ICU and 88 (12.6%) in non-ICU.

This is a retrospective case-series study, which included all deaths occurring in a private hospital, with open medical staff, between January 1, 2006 and July 31, 2010. Based on a protocol developed by the Committee for the Investigation of Hospital Deaths, a database was created with the minimum necessary information to outline the clinical profile of patients who died, as well as the time of occurrence of these deaths in the institution.

There was no difference in age between the patients who died in the ICU (80 ± 12 years) compared to non-ICU (78 ± 14 years) (p = 0.13), as there was no difference in the proportion of women between the ICU and non-ICU (44% vs. 49%, respectively, p = 0.56).

As a direct cause of death (Table 1) neoplasms were more prevalent in the non-ICU than in the ICU (45.5% vs. 7.0%, respectively, p = 0.0001), while sepsis was more prevalent in the ICU, although not statistically significant (65% vs. 23.7%, p = 0.07). All patients that died of sepsis at the non-ICU were receiving comfort measures. The prevalence of AMI as the direct cause of death was similar between the ICU and the non-ICU (2.8% vs. 1%, p = 0.61); the same was observed regarding cerebrovascular accident (5.2% vs. 5.7%, p = 1.0) and thromboembolism (1.0% vs. 1.1%, p = 1.0).

As for the underlying disease (Table 2), cardiovascular diseases were more prevalent in patients who died in the ICU than in non-ICU (46.0% vs. 17.0%, p = 0.0001), while neoplasias were more prevalent in the non-ICU (65.9% vs. 17.0%, p = 0.0001).

Infectious disease as the underlying cause of death was more prevalent in the ICU than in the non-ICU (30.8% vs. 11.4%, p = 0.001). On the other hand, there was no significant difference in the prevalence of dementia (2.4% and 4.5%, p = 0.45), kidney failure (10.4% vs. 4.5%, p = 0.11) and other diseases together (28% vs. 36.4%, p = 0.16) between the ICU and non-ICU, respectively. More patients in the ICU had more than one underlying cause of death than those admitted at the non-ICU (62% vs. 44%, respectively, p = 0.02).

Of the patients who died at the non-ICU, 68 (77.3%) received comfort measures, compared with 30 (14.2%) at the ICU (p = 0.0001).

When comparing the time of occurrence of deaths in the ICU and the non-ICU, it was observed that significantly more deaths occurred in the early hours of the day, between 6 am and 12 pm, in the first compared to the second (38% vs. 21%, p = 0.004).

When analyzing the period of occurrence of deaths in the ICU, we observed that 21% of the deaths occurred between 6 am and 12 pm, 30% between 12 pm and 6 pm, 26% between 6 pm and 12 am and 24% between 12 am and 6 am (p = 0.13) , whereas, at the non-ICU, 38% of the deaths occurred between 6 am and 12 pm, 18% between 12 pm and 6 pm, 19% between 6 pm and 12 am and 25% between 12 am and 6 am (p = 0.005).

Analyzing the occurrence of deaths every two hours, it can be observed that in the non-ICU there is an increase in the number of deaths from 4 am, peaking at around 8 am and with a small increase between 12 pm and 6 pm, i.e., those deaths follow a periodicity (Figure 1), whereas this pattern is not observed in the ICU (Figure 2).

Discussion

Our study shows that, when analyzing the occurrence of deaths in relation to the circadian pattern, there was no time of day associated with increased mortality in the ICU, i.e., the circadian rhythm in this unit mortality was for some reason disrupted (Figure 2). Also, there was no difference in the occurrence of deaths between the routine hours (6 am to 6 pm) and the shifts in the ICU (6 pm to 6 am), i.e., 51% versus 49% (p = 1.0). On the other hand, we observed a higher mortality rate in the non-ICU in the morning hours, when compared to other periods of the day, which configures the maintenance of the circadian pattern of mortality in this unit (Figure 2). Considering these results, we speculate that patients admitted to the non-ICU preserve the circadian periodicity, regardless of the underlying disease that affects them.

In the ICU, on the other hand, the role of several factors might change that frequency, either due to the confinement, the uninterrupted use of light throughout the day, use of vasoactive drugs, sedation, intubation, etc. To the best of our knowledge, this observation of the circadian pattern maintenance in the deaths occurring in the non-ICU had not been previously published. However, it coincides with the description by other authors of the occurrence of acute myocardial infarction and sudden death in the general population^6,7.

Additionally, Meynaar et al.⁸, when analyzing the influence of admission out of the routine schedule (shifts) on mortality in the ICU showed that admission at these times was not associated with increased mortality. In our study, although we did not analyze the association of the time of admission with mortality, we observed that there was no difference in mortality between the four analyzed periods of the day in the ICU.

As expected, the number of deaths in the ICU was higher than in the non-ICU, which is consistent to what has been observed in other services^14-16. Our study also shows that 77.3% of patients who died at the non-ICU were receiving comfort measures compared to 14% in the ICU.

These data are consistent with those reported in the literature^15,17. Analyzing our data within the context of terminal life, it is evident that neoplasias had a significant effect on the difference observed between the two units. Furthermore, all patients that died as a direct consequence of sepsis in the non-ICU were receiving comfort measures. It is important to report that the reason why the proportion of patients receiving comfort measures in the ICU was lower than that in the non-ICU, was that many of the deaths occurring in the ICU were due to multiple organ failure, therefore constituting an end-of-life process and the uselessness of maintaining therapeutic measures that would extend the vital status of patients.

Regarding sex, our study shows that there was prevalence of males in both units, that is, 56% in the ICU and 51% in the non-ICU, although not statistically significant. These data are different from those reported by Toscanini et al.¹⁸ in an analysis of deaths in various hospitals in Italy, which showed a predominance of female deaths in those institution wards. Moritz et al.¹⁵, in a study conducted in our country, observed the equivalence between the sexes in the deaths that occurred in a university hospital in the state of Santa Catarina, notwithstanding the predominance of male deaths in the ICU of that institution.

An unusual aspect of the results of our study was the prevalence of more than one underlying disease and cardiovascular diseases in patients who died in the ICU, when compared with those who died at the non-ICU, especially when there is no difference in age or sex between the patients from the two analyzed units. Assuming that 77% of patients who died at the non-ICU were receiving comfort measures as a result of the terminal stages of their illnesses, we imply that the presence of more than one underlying disease in these patients was as prevalent as that in patients who died in the ICU.

This fact is even more relevant when one considers that cancer and cardiovascular diseases have common risk factors, subjecting, at least theoretically, their carriers to the same possibilities of concomitant diseases¹⁹. Another possibility is the lack of information about concomitant diseases in the death certificate files that come from the non-ICU. According to Marins et al.²⁰, in 52.8% of all deaths in the city of Ribeirão Preto, there was no reference to secondary diagnoses in the death investigations.

When analyzing the report by Toscanini et al.¹⁸ regarding deaths that occurred in the wards of several hospitals in Italy, we found that, similarly to our study, there was a prevalence of malignancy as the underlying disease in patients that died in those institutions.

Limitations

Our study has several limitations. As records were not analyzed in their entirety, it is possible that other factors, which were not assessed, might have somehow modified the results. Due to this limitation we were unable to assess the type of medication that patients were using in their respective units at the hour, on the days or in the weeks before the deaths. Additionally, we did not assess the impact of therapeutic measures used, especially in the ICU on the circadian variation of death occurrence. We emphasize that our analysis was restricted to deaths, and thus, the clinical profile does not accurately depict the patients admitted to the studied units.

Conclusion

At the non-ICU, deaths occur more often in the morning period and follow a circadian rhythm, which does not occur at the ICU.

References

• 1. Panda S, Hogenesch JB, Kay SA. Circadian rhythms from flies to human. Nature 2002;417:329-35.
• 2. Brandstätter R. Encoding time of day and time of year by the avian circadian system. J Neuroendocrinol 2003;15:398-404.
• 3. Markus RP, Afeche SC, Barbosa Jr EM, Lotufo CMC, Ferreira ZS, Cipola-Neto J. Glândula pineal e melatonina. In: Marques N, MennaBarreto L. Cronobiologia: princípios e aplicações. São Paulo: Edusp; 2003. pp.191-222.
• 4. Menna-Barreto L. O tempo na biologia. In: Marques N, Menna- Barreto L. Cronobiologia: princípios e aplicações. São Paulo: Edusp; 2003. pp.26-9.
• 5. Paranjpe DA, Sharma VK. Evolution of temporal order in living organisms. J Circadian Rhythms 2005;3:7-19.
• 6. Muller JE, Stone PH, Turi ZG, Rutherford JD, Czeisler CA, Parker C et al. Circadian variation in the frequency of onset of acute myocardial infarction. N Engl J Med 1985;313:1315-22.
• 7. Cohen MC, Rohtla KM, Lavery CE, Muller JE, Mittleman MA. Meta-analysis of the morning excess of acute myocardial infarction and sudden cardiac death. Am J Cardiol 1997;79:1512-6.
• 8. Meynaar IA, Spoel JI, Rommes JH, Spreuwel-Verheijen M, BosmanRJ, Spronk PE. Off hour admission to an intensivist-led ICU is not associated with increased mortality. Critical Care 2009;13:1-7.
• 9. Arabi Y, Alshimemeri A, Taher S. Weekend and weeknight admissions have the same outcome of weekday admissions to an intensive care unit with onsite intensivist coverage. Crit Care Med 2006;34:605-11.
• 10. Barnett MJ, Kaboli PJ, Sirio CA, Rosenthal GE. Day of the week of intensive care admission and patient outcomes: a multisite regional evaluation. Med Care 2002;40:530-9.
• 11. Laupland KB, Shahpori R, Kirkpatrick AW, Stelfox HT. Hospital mortality among adults admitted to and discharged from intensive care on weekends and evenings. J Crit Care 2008;23:317-24.
• 12. Luyt CE, Combes A, Aegerter P, Guidet B, Trouillet JL, Gibert C et al Mortality among patients admitted to intensive care units during weekday day shifts compared with "off" hours. Crit Care Med 2007;35:3-11.
• 13. Antman E, Bassand JP, Klein W, Ohman M, Sendon JLL, Rydén L et al. Myocardial infarction redefined-a consensus document of The Joint European Society of Cardiology/American College of Cardiology committee for the redefinition of myocardial infarction: The Joint European Society of Cardiology/ American College of Cardiology Committee. J Am Coll Cardiol 2000;36:959-69.
• 14. Travassos C, Noronha JC, Martins M. Mortalidade hospitalar como indicador de qualidade. Ciênc Saúde Coletiva 1999;4:367-81.
• 15. Moritz RD, Beduschi G, Machado FO. Avaliação dos óbitos ocorridos no Hospital Universitário da Universidade Federal de Santa Catarina (HU/UFSC). Rev Assoc Med Bras 2008;54:390-5.
• 16. Ferrand E, Jabre P, Genod CV, Aubry R, Badet M, Badia P et al For the French Mort-a-l'Hôpital Group. Circumstances of death in hospitalized patients and nurses' perceptions. French Multicenter Mort-a-l'Hôpital Survey. Arch Intern Med 2008;168:867-75.
• 17. Goodlin SJ, Winzelberg GS, Teno JM, Whedon M, Lynn J. Death in the hospital. Arch Intern Med 1998;158:1570-2.
• 18. Toscani F, Di Giulio P, Brunelli C, Miccinesi G, Laquintana D. How people die in hospital general wards: a descriptive study. J Pain Symptom Manage 2005; 30:33-40.
• 19. 2008-2013 - Action Plan for the Global Strategy for the Prevention and Control of Noncommunicable Diseases. [citado 14 fev 2011]. Disponível em: http://www.who.int/cancer/prevention/en/
• 20. Martins M, Blais R, Leite IC. Mortalidade hospitalar e tempo de permanência: comparação entre hospitais públicos e privados na região de Ribeirão Preto, São Paulo, Brasil. Cad Saúde Pública 2004;20 (Supl 2):S268-S82.

Publication Dates

• Publication in this collection
  14 Oct 2011
• Date of issue
  Oct 2011

History

• Received
  21 Feb 2011
• Accepted
  11 July 2011