Reduced health-care utilization among people with chronic medical conditions during coronavirus disease 2019

Introduction

Coronavirus disease 2019 (COVID-19) is a major public-health crisis affecting nations across the world. ¹ Several studies have been conducted to assess the direct public-health impact of the COVID-19 outbreak. ² Of particular concern have been people with pre-existing chronic medical conditions who are at risk of experiencing high morbidity and mortality directly due to the COVID-19 pandemic. ³ However, the outbreak also has the potential to impact people with pre-existing chronic medical conditions indirectly if they are not able to access routine health care for managing their chronic conditions. Routine medical treatments for patients with chronic conditions are being delayed due to mobility restrictions, shortage of supplies of medicines and lack of human resources in clinics and hospitals.^4,5 In the early phase of the COVID-19 pandemic, many people may have perceived the risk of exposure to COVID-19 to be high in hospitals and clinics. This may have influenced their health-care-seeking behaviour for chronic medical conditions, thereby resulting in morbidity and mortality that cannot be directly attributed to the COVID-19 outbreak.

In this study, we aim to assess the extent, reasons and correlates of reduced health-care utilization among people with chronic medical conditions in Singapore during the COVID-19 outbreak. Studies show that the elderly are particularly vulnerable to the risk of complications and to dying due to COVID-19. ⁶ We hypothesized that the elderly (⩾60 years), people with a greater worry about contracting COVID-19 and those who perceive themselves to be at high risk of dying from COVID-19 would be more likely to miss their health-care appointments voluntarily.

Methods

Through a market research company, we administered a one-time web survey to its panel members between 31 March 2020 and 14 April 2020. We quota sampled based on Singapore’s census distribution for age, sex and ethnicity, with a ±5% error margin to ensure national representativeness. Participants included Singapore citizens, permanent residents and long-term pass holders (i.e. those with an employment pass, work pass, dependant’s pass or student pass). All participants provided written informed consent and answered the survey in English. Those who completed the survey received small incentives via a point-based reward system from the survey company.

Participants self-reported their chronic medical conditions. For each condition, we asked participants whether they had missed a health-care appointment since the ‘Disease Outbreak Response System Condition’ (DORSCON), a colour-coded framework for infectious disease outbreaks in Singapore, turned Orange (7 February 2020), signifying that disease is severe and spreading easily. Participants reported their reasons for missing their appointment(s), including if it was cancelled voluntarily (demand driven) or by their provider (supply driven). Participants were allowed to select multiple reasons for cancelling their appointment. We assessed participants’ risk perceptions regarding the severity of COVID-19. First, participants reported their degree of worry about the risk of contracting COVID-19 on a five-item scale, ranging from ‘not worried at all’ to ‘very worried’. Additionally, participants also responded to the likelihood of dying upon contracting COVID-19 on a four-item Likert scale, ranging from ‘very unlikely’ to ‘very likely’ (Appendix Table A1).

For the sub-sample of participants with chronic conditions who cancelled their health-care appointment voluntarily, we used a binomial logistic regression (controlled for sex) to test our hypotheses. We used Stata v16.1 (StataCorp, College Station, TX) for analyses.

Results

Of the 1338 participants who received the survey, 98 did not respond, and 223 did not complete the survey. The 1017 participants who completed the survey (response rate: 76%) had a median age of 40 years. About half (47.6%) were females, and the majority (84.2%) were Chinese. Overall, our sample resembled the distribution in the general population in terms of sex (51% female) and age (median=41.1 years) but over-represented people of Chinese ethnicity (74.3%). ⁷ Of these 1017 participants, 349 (34%) reported a chronic medical condition. The median age of these participants was 48 years, and less than a quarter of the participants (22.9%) were elderly (⩾60 years; Table 1). About 23% of the participants with chronic conditions reported they were very worried about contracting COVID-19, and nearly one tenth (8.9%) of the participants perceived that they were ‘very likely’ to die upon contracting COVID-19. About 40% (n=136) of the participants reported having missed at least one health-care appointment in the last two months (M=57.4 days; SD=2.6 days) since the DORSCON level in Singapore changed to orange. Of these, 39% (n=53) reported that their provider had cancelled the appointment, and 72% (n=98) reported having cancelled at least one appointment voluntarily. Just over half (52%; n=71) of the appointments were scheduled at the hospital specialist outpatient clinics, and 50% (n=68) of the appointments were scheduled at the polyclinics. More than 90% of the patients with chronic kidney disease reported having missed their health-care appointment.

OPEN IN VIEWER

Table 1.

Sample characteristics of participants with a chronic medical condition and their health-care utilization (N=349).

Demographic characteristics
Age, median (IQR)	48 (37–58)
Age, categorical (years), n (%)
<45	150 (43.0)
45–59	119 (34.1)
⩾60	80 (22.9)
Female, n (%)	151 (43.3)
Education, n (%)
Below university	161 (46.1)
University and above	188 (53.9)
Number of chronic conditions, median (IQR)	2 (1–3)
Perceived risk for COVID-19
Worry about contracting COVID-19, n (%)
Not worried/slightly worried	74 (21.2)
Somewhat worried	93 (26.6)
Quite worried	101 (28.9)
Very worried	81 (23.2)
Likelihood of death upon contracting COVID-19, n (%)
Very unlikely	69 (19.8)
Somewhat unlikely	139 (39.8)
Somewhat likely	110 (31.5)
Very likely	31 (8.9)
Missed/cancelled health-care appointment, n (%)	136 (39.0)
Reason for missing/cancelling health-care appointment, a n (%)	N=136
Health-care provider cancelled	53 (39.3)
Participant cancelled because of at least one of the reasons b	98 (72.1)
Afraid of contracting COVID	72 (52.9)
Afraid of transmitting COVID to family members	64 (47.1)
Was not feeling well enough to attend the appointment	25 (18.4)
Location of missed/cancelled appointment, a n (%)
Hospital inpatient ward	37 (27.2)
Hospital specialist outpatient clinic	68 (50.0)
General physician clinic	43 (31.6)
Polyclinic	71 (52.2)
Home medical/nursing care	17 (12.5)
Missed/cancelled health-care appointment by type of chronic medical condition	Yes
High cholesterol, N=145	63 (43.4)
High blood pressure, N=119	45 (37.8)
Heart disease, N=77	42 (54.5)
Chronic joint pain, N=68	33 (48.5)
Chronic back pain, N=62	39 (62.9)
Diabetes, N=52	26 (50.0)
Chronic gastrointestinal disease, N=42	26 (61.9)
Chronic lung condition, N=32	17 (53.1)
Osteoporosis, N=32	24 (75.0)
Chronic eye condition, N=34	27 (79.4)
Neurological condition, N=21	13 (61.9)
Cancer, N=18	12 (67.0)
Chronic kidney disease, N=12	11 (91.7)

a

Numbers and percentage do not add up to total, as more than one option could be selected.

b

Numbers and percentage do not add up to total, since more than one individual specific reason could be selected.

IQR: interquartile range; COVID-19: coronavirus disease 2019.

Results from logistic regression showed that contrary to our hypothesis, elderly (⩾60 years) (odds ratio (OR)=0.43, 95% confidence interval (CI) 0.24–0.77; p<0.01) and middle-aged (45–59 years) participants (OR=0.32, 95% CI 0.15–0.65; p<0.01) were less likely than younger adults (<45 years; reference) to cancel their health-care appointment voluntarily. As hypothesized, the odds of participants missing their appointment voluntarily increased with their worry about contracting COVID-19 and their perceived risk of dying due to COVID-19. Participants who were very worried about contracting COVID-19 compared to those that were not worried or slightly worried (reference) were more likely to cancel their health-care appointment voluntarily (OR=3.21, 95% CI 1.48–6.97; p<0.01). Similarly, participants who perceived themselves as ‘very likely’ to die upon contracting COVID-19 compared to ‘very unlikely’ (reference) group were more likely to cancel their health-care appointment voluntarily (OR=2.78, 95% CI 1.03–7.51; p<0.05).

Discussion

To reallocate health-care resources towards managing COVID-19 patients and to avoid cross-infections among patients, health-care providers worldwide have reduced outpatient visits, admissions and elective surgeries. ⁸ Whereas cancellation of appointments by health-care providers has reduced health-care use among people with chronic conditions, it is notable that most people with a chronic condition reported cancelling their appointments voluntarily. The elderly (compared to younger adults) may have perceived the risk from cancelling their appointment to be greater than the risk of contracting COVID-19, and were thus less likely to cancel voluntarily. On the other hand, those with a greater worry of contracting COVID-19 and a perceived risk of death due to COVID-19 were more likely to cancel their appointments voluntarily.

While chronic conditions such as high blood pressure or diabetes can be managed through lifestyle changes and continuation of prescribed medications, missed health-care appointments for conditions such as chronic kidney disease, cancer or heart disease may have serious implications. It can increase the risk of worsening health status and of mortality that would otherwise not directly be attributable to COVID-19. This indirect health impact of the outbreak may last long after the outbreak has been controlled.

These findings have implications for making telehealth widely and urgently available within national health-care systems in order to allow continuity of care for people with chronic medical conditions and to allow triaging of those who need more intense management.

A limitation of our study was that we conducted it with a convenience sample of adults, which may limit generalizability. However, we believe that the phenomenon of reduced health-care utilization is likely to be widespread across countries affected by the COVID-19 outbreak. Our results are also supported by studies from China and Italy that report reduced hospitalization rates among patients with cardiovascular illnesses, as many patients with acute deterioration in their illnesses have avoided going to hospital.^5,6,9 Another limitation was that the use of a web-survey restricted participation to those who are Internet savvy and who speak English. However, conducting a web-survey enabled us to collect time-sensitive information faster than other methods of data collection. We also did not have information on whether appointments missed were at a public or a private hospital or clinic. Lastly, it is possible that participants who reported a missed health-care appointment subsequently went to another health-care provider or had sufficient supply of medications, and therefore management of their chronic condition was not disrupted. We did not capture these in our survey.

In conclusion, we find reduced health-care utilization among people with a chronic medical condition that is both supply and demand driven. Urgent efforts are required to ensure continuity of care for people with chronic medical conditions in order to avoid any long-term impact on their health and mortality.

Supplemental Material