The impact of dementia on drug costs in older people: results from the SNAC study

Study population

The Swedish National Study on Aging and Care (SNAC) is an ongoing, population based,
longitudinal study of aging and health conducted at four different sites in Sweden.
We analyzed data from the baseline examination conducted in 2001–2004 from Nordanstig
in the middle part of Sweden and from Kungsholmen/Essingeöarna in the central part
of Stockholm. Inclusion criteria were having an address in either of the actual areas
at time of birthday for the ages specified below.

The SNAC study has been described in detail elsewhere 16]. In short, people aged 60, 66, 72, 78, 81, 84, 87 and ?90 years are interviewed by
a nurse about a wide range of domains including socioeconomic status, living habits
and family history. Participants are also examined by a physician, memory tested by
a psychologist and laboratory tests are collected. Data about diseases and drug use
are collected during the interview with the physician. When the participant is not
able to provide information, a relative is asked instead. If the person lives in an
institution, the information is most often collected from medical records and staff.

The care system for older people in Sweden

In Sweden, care for the elderly – and the associated costs – are divided between municipalities
and the county council. Social care (e.g. home services, long term institutional care
and day care) is covered by the municipalities while primary health care and specialist
care are organized by county councils. Individual drug expenditure is to a great extent
subsidized in Sweden. In 2003, the maximum level of out of pocket expense for drugs
was 1,800 SEK per 12 month period. Overall, the majority of costs for social and medical
care in Sweden are publically funded by taxes.

Definitions

Socio-demographic variables

Age was categorized into 60–69, 70–79, 80–89 and ?90 years in the descriptive analysis
and used as a continuous variable in the Generalized Linear Model (GLM). Residential
setting was dichotomized into community-dwelling (i.e. living in one’s own home) vs.
institution (sheltered accommodation, old people’s home, group dwelling or nursing
home). Educational level was classified into elementary vs. additional schooling.

Dementia diagnosis

Dementia was diagnosed by physicians according to Diagnostic and Statistical Manual
of Mental Disorders DSM III-R 17].

Drugs

Drug data were self-reported (or reported by a proxy) 10]. Participants were asked to bring a list of currently used drugs to the interview.
Prescriptions and medical containers were also inspected when available. Drugs were
classified according to their Anatomical Therapeutic Chemical (ATC) code, as recommended
by the World Health Organization (WHO) 18].

Drug costs were calculated based on a register with drug prices from the National
Corporation of Swedish Pharmacies (Apoteket AB) from 2003. Each drug of each subject
was sought out in the drug register. Thereafter, a matching preparation and strength
was looked up. Among the records in the register that fulfilled these criteria a suitable
package was selected. For tablets or capsules, packages with 100 or close to 100 tablets/capsules
were selected. For other preparations, such as mixtures, the largest package was selected.
The price of the package was divided by the number of units; for example number of
tablets/capsules, or number of ml for liquid preparations. The obtained price per
unit was then multiplied with the number of units taken daily by the subject. For
drugs taken as needed we instead calculated the price per Defined Daily Dose (DDD),
which is the average daily dose of a drug when used for its main indication in an
average 70 kg adult, as established by WHO, and we assumed that as needed drugs were
taken in an average dose of half a DDD per day. For anti-infective drugs, we assumed
a limited treatment period of 20 days/year 15]. The costs are expressed in the Swedish currency SEK. In December 2003, 100 SEK corresponded
to 13.27 USD. The results are mainly presented as costs per individual, which means
the total cost for a drug or drug group in a group of elderly people, divided by the
total number of subjects in that group. If the cost is instead presented per user,
then that is noted in the text.

Comorbidity

A modified Charlson Comorbidity Index was used to adjust for confounding by co-morbidities
19], 20]. The index contains seven diagnoses with the weight of one (myocardial infarction,
congestive heart failure, cerebrovascular disease, dementia, chronic pulmonary disease,
connective tissue disorder and diabetes without complication) and two diagnoses with
the weight of two (moderate or severe renal failure and any tumour), which gives a
total of maximum eleven. The comorbidity index was entered as a categorical variable
in the analysis (none, one, two and three or more modified Charlson index weights).
All diagnoses except renal failure and dementia diagnosis were based on medical history
and examination of the physician. Renal failure was calculated from Cockgroft-Gaults
formula 21] and defined as creatinine clearance??25 ml/min. The dementia diagnosis procedure
is described above.

Physical functioning

Physical functioning was assessed by the Katz index of Activities of Daily Living
(ADL) 22], which is a scale that measures dependency in six basic activities: bathing, dressing,
going to the toilet, transferring, feeding and continence. Functional independence
was defined as no need of assistance, partially dependent was defined as being dependent
in one or two activities and being dependent was defined as dependent in three or
more activities.

Statistical analysis

The cost data had a non-normal and skewed distribution. Accordingly, regression analysis
of costs was performed by using a generalized linear model (GLM) with the assumption
of a gamma shaped distribution of the dependent variable 23]. GLMs are generally well suited for statistical analysis of cost data, which often
show a high degree of non-normality.

We used a two-step procedure. First, logistic regression with costs as binary outcome
was performed in order to understand which of the factors that was associated with
the highest costs. Second, a GLM model was run to explore the magnitude of the cost-driving
factors. In the GLM, the cost driving factors were dichotomized and first entered
separately. All models were adjusted for age, gender and education. In the joint analysis,
all factors were entered simultaneously. Dementia diagnosis is included in the Charlson
Comorbidity Index, but in the joint analysis, dementia was analyzed as a separate
variable and, thus, was removed from the index. Results of the GLM are expressed as
relative change in cost due to increase in the explanatory variable (e.g. dementia).
Inclusion of site (i.e. Kungsholmen and Nordanstig) did not affect the results; therefore,
this variable was not included in the analyses. IBM SPSS Statistics 22.0 for Windows
(IBM corp., New York, NY, USA) was used for the analyses.

Ethics

The study was approved by the ethical review board in Stockholm (dnr 01–114) and in
Uppsala (dnr 01–123). All participants were informed about the study and thereafter
gave a written informed consent, and if not possible a proxy (i.e. a spouse or next
of kin) gave the consent.