Skip to main content

Table 4 Logistic regression of use (vs. no usea) of healthcare in BC residents, 2015/16b

From: Population segments as a tool for health care performance reporting: an exploratory study in the Canadian province of British Columbia

Odds Ratio (OR) (LCL – UCL)

 

Segment 1 Low need (n = 2,558,276)

Segment 2 Multiple morbidities (n = 449,925)

Segment 3 Medically complex (n = 116,821)

Segment 4 Frail (n = 65,661)

Age (years)

 18–44

0.52 (0.51–0.53)

0.42 (0.36–0.49)

0.57 (0.46–0.71)

n/a

 45–64

0.63 (0.61–0.64)

0.69 (0.61–0.77)

0.83* (0.67–1.04)

n/a

 65–74

ref

ref

ref

ref

 75+

1.19 (1.15–1.23)

0.84 (0.74–0.96)

0.83* (0.63–1.09)

1.02* (0.82–1.26)

Sex

 Female

2.13 (2.11–2.15)

1.67 (1.53–1.83)

2.08 (1.84–2.35)

1.16* (0.97–1.40)

 Male

ref

ref

ref

ref

Number of chronic conditions (0–5+): continuous variablec

4.67 (4.60–4.74)

2.05 (1.87–2.25)

1.76 (1.63–1.90)

1.45 (1.36–1.55)

SES

 Low

0.95 (0.94–0.96)

0.77 (0.70–0.84)

0.87 (0.77–0.98)

0.93* (0.78–1.11)

 High

ref

ref

ref

ref

  1. *p ≥ 0.05, all other p < 0.05, LCL lower confidence limit, UCL upper confidence limit; SES socioeconomic status (low SES income quintiles 1, 2, high SES income quintiles 3, 4, 5.)
  2. aNo health care use is defined as $0 in health care costs in 2015/16.The number of individuals with $0 in health care costs in 2015/16 varies by segment: Segment 1 = 264,375; Segment 2 = 2035; Segment 3 = 1095; Segment 4 = 505
  3. bNote that this table excludes individuals with no FP visits in the 3 years of data used to create the continuity of care measure (UPC). The number of people excluded varies by segment: Segment 1 = 249,449; Segment 2 = 272; Segment 3 = 815; Segment 4 = 174
  4. cNumber of chronic conditions was treated as a continuous variable given that the number of chronic conditions varies by segment (e.g., by definition, segment 1 has fewer chronic conditions than segment 4); please see Supplementary File 3 (Table 1a and b) for analyses where chronic conditions were treated as categorical variables; we note that this did not change our findings