JMIR Publications

ABSTRACT

Background:

Healthcare avoidance in the COVID-19 Pandemic has been widely reported. Yet few studies have investigated the dynamics of hospital avoidance behaviour during pandemic waves and inferred its impact on excess non-COVID-19 death toll.

Objective:

To measure the impact of hospital avoidance behaviour on excess mortality using emergency department (ED) patient data from 2016 to 2021, during which Hong Kong experienced a unique COVID-19 pandemic with four distinct waves of case number surges.

Methods:

Our data is taken from the CDARS Hong Kong Hospital Authority administrative database, which oversees all local public hospitals and plays a prominent role in emergency care provision. To estimate excess mortality, two-stage least squares was utilised with daily tallies of ED visit and 28-day mortality. Elderly records were categorised by the residential care home for elderly status (RCHE) and comorbidities were used to explain the demographic and clinical attributes of excess 28-day mortality.

Results:

Compared with the average in 2016-2019 average there was a reduction in total ED visits in 2020 of 25·4%. During the same period, the 28-day mortality of non-COVID-19 ED deaths increased by 7·82% compared with 2016-2019. The estimated total elderly excess non-COVID 28-day death by reduced ED visits throughout 2020 to 2021 is 1,958 (1,100-2,820, no time lag). The actual excess death in 2020 and 2021 are 3,143 and 4,013 respectively, with 2016-2019 average as the benchmark. Death on Arrival (DOA)/ Death before Arrival (DBA) increased by 35·1% in 2020, while non-DOA/DBA mortalities increased only by a moderate 4·65%. In both DOA/DBA and non-DOA/DBA, the increases were higher during wave periods than in non-wave periods. Moreover, non-RCHE patients saw a greater reduction in ED visit than RCHE residents across all waves by more than 10%. Most of the subset comorbidities demonstrated an annualised reduction in visit in 2020. Renal diseases and severe liver diseases saw a notable death increase.

Conclusions:

We demonstrated a statistical method to estimate hospital avoidance behaviour during a pandemic, and quantified the consequential excess 28-day mortality, with a focus on elderlies, who had high frequencies of ED visit and deaths. This study serves as an informed alert and possible investigation guideline to healthcare professionals about hospital avoidance behaviour and its consequences.