Mortality and length of stay are highest in heart failure patients admitted in January, on Friday, and overnight, according to research presented at the Heart Failure Congress 2013 in Lisbon, Portugal (25–28 May). The analysis of nearly one million heart failure admissions over 14 years was presented by David P Kao, Denver, Colorado.
Identifying peaks in admissions and mortality should assist targeted resource allocation at higher risk times. Seasonal, weekly and hourly variations have been observed in heart failure admissions but the reasons are unclear. Until now, the relationship of these variations with mortality and length of stay has not been investigated in a single study.
The current study examined the impact of day, month and hour of admission on in-hospital mortality and length of stay in 949,907 hospitalisations for congestive heart failure. Data was analysed from all hospitals in the state of New York from 1994 to 2007. A greater number of factors were included in the analysis than ever before so that the researchers could confirm or refute previous theories on the reasons behind variations in heart failure morbidity and mortality (ie. substance use).
The researchers found that daily heart failure admissions increased significantly over this period (+1.1 admissions/day/year) while in-hospital mortality and length of stay decreased (-0.3%/year and -0.3 days/year, p<0.0001 for all).
“These findings confirm the huge decline in mortality in hospitals for heart failure over the past 14–15 years following major advances in therapy,” said Kao.
Daily heart failure admissions peaked in February (p<0.0001), while in-hospital mortality (p<0.0001) and length of stay (p=0.01) peaked in January. Mortality and length of stay were lowest for admissions between 06h00-12h00 and highest overnight (18h00-24h00) by a small margin (adjusted OR of death 1.22, p<0.0001). Mortality and length of stay were lowest in patients admitted on Monday (adjusted OR of death 1.09, p<0.001) and highest on Friday (p<0.0001).
Numerous theories have been mooted for the cause of seasonal variations in heart failure morbidity and mortality, for example, that the holiday spike is caused by alcohol and drug use.
“For the first time, we have shown that there was not a higher rate of alcohol and drug use reported in heart failure patients during December and January, when heart failure mortality was the highest,” said Kao.
Seasonal variations affected rate of heart failure hospitalisation and mortality in patients over the age of 30, and the effect was greater with advancing age. An increase in concurrent pneumonia in the winter could impact on heart failure mortality, but there was less seasonal variation in other respiratory diseases like chronic obstructive pulmonary disease (COPD).
The findings suggest that staffing may have an impact on seasonal variations in mortality and length of stay. Kao said: “The fact that patients admitted right before the weekend and in the middle of the night do worse and are in hospital longer suggests that staffing levels may contribute to the findings.”
He added: “The seasonal effect on in-hospital death from heart failure remained even after controlling for time and weekday of admission, 17 other medical conditions including substance use, kidney disease, and pneumonia, and demographic factors including gender, ethnicity, and medical coverage status. Seasonal variations in morbidity and mortality occur in many diseases, particularly heart disease, and the cold weather itself may have a part to play.”
“Doctors and hospitals need to be more vigilant during these higher risk times and ensure that adequate resources are in place to cope with demand. Patients should be aware that their disease is not the same over the course of the year and they may be at higher risk during the winter. People often avoid coming into hospital during the holidays because of family pressures and a personal desire to stay at home but they may be putting themselves in danger,” concluded Kao.