Healthcare use in the year preceding a diagnosis of pancreatic cancer: a register-based cohort study in Denmark

Figures & data

Table 1. Study population characteristics of pancreatic cancer patients diagnosed in 2012–2018 and references without pancreatic cancer^a.

	Pancreatic cancer patients	Reference population
	N (%)	N (%)
Sex
Male	3103 (52.4)	31,030 (52.4)
Female	2823 (47.6)	28,230 (47.6)
Age at index date
Median, IQI^b	70 (64; 77)	70 (64; 77)
Age groups (years)
18–49	189 (3.2)	1905 (3.2)
50–64	1412 (23.8)	14,068 (23.7)
65–74	2405 (40.6)	24,037 (40.6)
75–84	1575 (26.6)	15,832 (26.7)
85+	345 (5.8)	3418 (5.8)
Educational level
Basic	2286 (38.6)	22,052 (37.2)
Short	2692 (45.4)	26,327 (44.4)
Long	948 (16.0)	10,881 (18.4)
Disposable income
Low	2073 (35.0)	18,873 (31.8)
Middle	1986 (33.5)	19,544 (33.0)
High	1867 (31.5)	20,843 (35.2)
Marital status
Cohabitant	3447 (58.2)	36,402 (61.4)
Living alone	2479 (41.8)	22,858 (38.6)
Country of origin
Danish	5627 (95.0)	56,067 (94.6)
Immigrant	299 (5.0)	3193 (5.4)
Comorbidity^c
Low	3541 (59.8)	38,129 (64.3)
Moderate	1675 (28.3)	15,645 (26.4)
Severe	710 (12.0)	5486 (9.3)
Diabetes^e
Yes	836 (11.1)	5137 (6.8)
Tumour stage
Localised	633 (10.7)	n/a^e
Regional	1031 (17.4)	n/a
Distant	3176 (53.6)	n/a
Unknown	108 (1.8)	n/a
Missing data	978 (16.5)	n/a
Surgery
Curative resection	1335 (22.5)	n/a
Palliative surgery	329 (5.6)	n/a
No surgery	4262 (71.9)	n/a

Significant differences between groups are shown in bold (Pearson’s chi-square test).

^a The references were matched 1:10 on sex, age and general practice. Matching on age was based on a maximum deviation of two years compared to the pancreatic cancer patients.

^b IQI: interquartile interval.

^c Charlson’s Comorbidity Index score defined at study entry: low = score 0, moderate = scores 1–2, severe = scores ≥3.

^d Defined at study entry.

^e n/a: not applicable.

Figure 1. Contacts* in general practice for 5926 pancreatic cancer patients and 59,260 references without pancreatic cancer. *Number of contacts include daytime face-to-face contacts, email consultations and telephone consultations in the 12 months prior to the pancreatic cancer diagnosis date and a corresponding index date assigned to references without pancreatic cancer. Number of contacts are presented as crude rates of mean number of contacts per month (upper part) and incidence rate ratios (lower part) adjusted for age, marital status, ethnicity, educational level, household income and comorbidity. Black lines represent 95% confidence intervals.

Figure 2. Tests* in general practice for 5926 pancreatic cancer patients and 59,260 references without pancreatic cancer. *Number of tests in the 12 months prior to the pancreatic cancer diagnosis date and a corresponding index date assigned to references without pancreatic cancer. Number of tests are presented as crude rates of mean number of tests per month (upper part) and incidence rate ratios (lower part) adjusted for age, marital status, ethnicity, educational level, household income and comorbidity. Black lines represent 95% confidence intervals.

Figure 3. Number of hospital contacts* for 5926 pancreatic cancer patients and 59,260 references without pancreatic cancer. *Number of contacts in the 12 months prior to the pancreatic cancer diagnosis date and a corresponding index date assigned to references without pancreatic cancer. Number of contacts are presented as crude rates of mean number of tests per month (upper part) and incidence rate ratios (lower part) adjusted for age, marital status, ethnicity, educational level, household income and comorbidity. Black lines represent 95% confidence intervals.

Figure 4. Number of diagnostic investigations* for 5926 pancreatic cancer patients and 59,260 references without pancreatic cancer. (A) All MR scans, CT scans, ultrasound and X-ray. (B) Ultrasound abdomen, MR thorax/abdomen/pelvis, CT thorax/abdomen/pelvis. *Number of diagnostic investigations in the 12 months prior to the pancreatic cancer diagnosis date and a corresponding index date assigned to references without pancreatic cancer. Number of diagnostic investigations are presented as crude rates of mean number of investigations per month (upper part) and incidence rate ratios (lower part) adjusted for age, marital status, ethnicity, educational level, household income and comorbidity in model A and comorbidity in model B. Black lines represent 95% confidence intervals.