Skip to main content
. 2010 Mar 15;8(3):147–161. doi: 10.4321/s1886-36552010000300001

Table 2.

Finnish research into factors associated with compliance, adherence and persistence in English language peer-reviewed journals since 2000

Author, date and reference number Study design and/or data source Study sample Relevant outcome measure(s) Main findings
Foulon et al. 201059 Register-based study (National Prescription Register) Children and adolescents aged ≤ 19 (n=27,676) who received a reimbursed antidepressant from 1998-2005 Percentage of children and adolescents dispensed only one antidepressant prescription (maximum duration 3 months) (S) 31% of children and adolescents received only one reimbursed antidepressant prescription (maximum persistence 3 months)
Lunnela et al. 201063 Qualitative interviews 12 compliant patients with glaucoma Perceived importance of education and social support received from healthcare personnel, relatives and peers among compliant patients with glaucoma Individual variation in the need for education and support in relation to compliance. Important to receive education and support during initial stages of illness and following treatment change
Helin-Salmivaara et al. 201040 Register-based study (National Prescription Register) 490,024 new users of statins in yearly cohorts from 1995 to 2005 One year persistence with statin therapy. Proportion of days covered among patients who discontinued within 365 days since initiation Compared with the 1995 cohort, the 1998, 1999, 2000, 2001, 2002, 2003, and 2004 cohorts were less likely to discontinue treatment within 12-months. Hazard ratios for discontinuation ranged from 0.91 (95% CI, 0.88-0.94) in 1998 to 0.80 (95% CI, 0.78-0.83) in 2004
Meretoja et al. 200950 Register linkage study (National Prescription Register, National Hospital Discharge Register) 5,586 patients discharged from hospital following admission for a stroke in 2005 Reimbursed purchases for secondary preventative (warfarin, aspirin+dipyridamole or clopidogrel) and cardiovascular medications in pharmacies pre and post hospital admission for stroke 33% of patients did not collect prescription(s) for any secondary preventative medication following hospital discharge. The prevalence of antihypertensive use increased from 64% pre-stroke to 78% post-stroke. The prevalence of statin use increased from 25% pre-stroke to 58% post-stroke
Tacke et al. 200951 Observational study 12 patients who received buprenorphine/naloxone for 6 days in a compliance-monitoring device Regularity of daily administration, treatment costs, and patients’ views on acceptability and drug diversion Six patients demonstrated good compliance and 2 patients demonstrated minor irregularities. Seven patients reported the package was too large, 5 patients positive treatment influence, 3 patients that it stopped diversion and 1 patient thought the package was difficult to use
Keränen et al. 200952 Retrospective medical record review 94 patients (mean age 59 years) with Parkinson’s Disease Persistence with pramipexole at 1, 3 and 6 years based on physicians records at university hospital 76%, 63% and 63% of patients were still taking pramipexole at 1, 3 and 6 years respectively. Predictors of discontinuation were orthostatic hypotension and entacapone treatment prior to initiation of pramipexole
Helin-Salmivaara et al. 200939 Register-based study (National Prescription Register) 562,598 new users of statins in Finland from 1995-2005 Discontinuation defined using period covered by the last prescription followed by a tablet-free gap exceeding 270 days Persons with diabetes (15%) were more likely to be taking a statin one-year after initiation compared to those without diabetes (77.1% vs. 73.8%). Risk for discontinuation was lower among male than female diabetic persons OR 0.94 (95%CI 0.91-0.96)
Pohjanoksa-Mäntylä et al. 200960 Qualitative interviews 29 people with depression who had used the Internet to access drug information Self-reported impact of using online drug information on medication taking behaviour (S) Self-reported impacts of utilising online drug information included being reassured, deciding to change a dose, to discontinue a medication and to suggest a new medication to a physician
Nabi et al. 200841 Register-based study (National Prescription Register) combined with postal survey 1,021 hypertensive patients aged 26-63 years participating in the Finnish Public Sector Cohort Study Treatment adherence defined according to 12-month reimbursed medication supply (0 days non-adherence, 1-364 days partial adherence, 365 days adherence). Sense of coherence measured using Orientation to Life Questionnaire 60% of patients were adherent, 36% partially adherent, and 4% nonadherent. A high sense of coherence was associated with lower odds of being nonadherent (OR 0.55; 95% CI, 0.31-0.96)
Holma et al. 200853 Prospective cohort study (Vantaa Depression Study) 218 patients with major depressive disorder Antidepressant adherence categorised as regular, somewhat irregular, very irregular, or not at all Antidepressant adherence in acute treatment phase independently predicted maintenance treatment (OR 3.18; 95% CI 1-12-9.03)
Helin-Salmivaara et al. 200838 Register-based study (National Prescription Register) 18,072 new users of statins in 1995 Persistence (time from initiation to discontinuation) and adherence (proportion of days covered by treatment) 10-year statin persistence in general population was 44%. Persons aged 45-74 years at initiation and those with ≥1 prescription for another cardiovascular drug most likely to continue to fifth year. Adherence was highest among persistent users (94%)
Sokero et al. 200858 Prospective cohort study (Vantaa Depression Study) 218 patients with major depressive disorder (mean age 41 years, 91 not suicidal, 92 suicidal ideation, 34 history of suicide attempts). Antidepressant adherence categorised as regular, somewhat irregular, very irregular, or not at all on basis of patient interview Self-reported adherence did not differ between those patients who were not suicidal (69%), had suicidal ideation (71%) or had a history of suicide attempts (70%)
Salomaa et al. 200747 Register linkage study (cause of death register, hospital discharge register, National Prescription Register) All patients aged 35-74 years hospitalized for the first nonfatal acute coronary syndrome in Finland during 1995-2003 (n=53 353) Medication persistence at 3 months assessed using National Prescription Register 28% and 15% of the patients did not receive hypolipidemic medications or beta-blockers after coronary syndrome and a further 6% and 10% discontinued these medications 3 months later
Pitkälä et al. 200742 Prospective study among participants in randomised controlled trial 400 cardiovascular patients (>74 years) Institutionalisation or death at 4.5 year follow-up Skepticism towards medications (HR 2.73, 95% CI 1.11-6.52) and non-adherence (HR 6.24, 95% CI 1.88-20.67) associated with institutionalisation or death at 4.5 year follow-up
Kyngäs et al. 200764 Cross-sectional postal survey using health register as a sampling frame 300 children and adolescents aged 13-17 years with insulin dependent diabetes (97% response rate) Adherence (including medication adherence) coded as a dichotomous variable Threat to mental wellbeing OR 7.68 (95% CI 3.95-13.42), support from physicians OR 6.69 (95% CI 3.70-11.46), support from nurses OR 6.28 (95% CI 2.78-16.25) and motivation OR 5.52 (95% CI 3.70-11.46) were factors that significantly predicted adherence
Melartin et al. 200554 Prospective cohort study (Vantaa Depression Study) 269 patients with DSM-IV major depressive disorder Adherence and attitudes toward antidepressants among 198 patients followed for 18 months 88% patients received antidepressants in acute phase, 49% terminated treatment prematurely. Termination was associated with major depressive episodes explained by fear of dependence and antidepressant side-effects
Lahdenperä et al. 200349 Development of a compliance scale using a self-completed patient survey A convenience sample of 103 patients with hypertension aged 24-65 years from 5 health centres Principal components analysis of compliance survey Data supported five subscales labelled lifestyle, intention, attitudes, responsibility and smoking. There was a correlation between medication use and attitudes (r=0.282, p<0.015). However, the two medication-related items were excluded from final scale because 27% of patients did not take medication.
Jokisalo et al. 200344 Cross-sectional mail surveys and health examination of general practice patients 1782 (80%) hypertensive patients from 26 health centres during one week in 1996 Factors associated with poor blood pressure control (BP≥ 140/90 mmHg) Non-compliance in men was associated with poor blood pressure control. Hopelessness towards hypertension (OR, 2.16; 95% CI 1.20-3.88), medium and high levels of frustration with treatment (OR, 1.50; 95% CI 1.04-2.18 and OR 1.83; 95% CI 0.98-3.44) were associated with poor control. Older age and monotherapy were also associated with poor blood pressure control.
Jokisalo et al. 200245 Cross-sectional mail surveys and health examination of general practice patients 1782 (80%) hypertensive patients from 26 health centres during one week in 1996, of whom 1561 on antihypertensive medication Self-reported non-compliance with antihypertensive medication (took medication ‘less often than prescribed during the last year’ or ‘modification of dosage instructions’) Perceived health system-related problems (OR 4.77, 95% CI 2.76-8.26) and patient-related problems (OR 3.23, 95% CI 1.79-5.81) associated with self-reported non-compliance. Adverse drug reactions also associated with non-compliance (OR 1.41, 95% CI 1.03-1.94)
Kampman et al. 200255 Prospective interview study 59 patients with a first-onset psychosis during their initial phase of treatment Observed compliance during the first 3 months of treatment assessed using patient record data Determinants of observed non-compliance included harmful side effects, male sex, lack of social activities, low PANSS score for positive symptoms, high PANSS total score and young age. The duration of untreated psychosis was not associated with compliance
Närhi et al. 200221 One-year intervention study in four community pharmacies 31 patients (aged 20-64 years) with perceived problems in asthma management Perceived problems in asthma self-management at baseline, 4 months, 8 months and 12 months Of 28 patients who completed the intervention, 17 patients self-reported problems with medications at baseline (including difficulty remembering to take medications, and considering medications unpleasant) compared to 4 patients at the 12-month follow-up
Kneckt et al. 200168 Cross-sectional patient self-completed survey 149 patients with insulin dependent diabetes mellitus Adherence to insulin regimen (adjusting injections to meal times and adjusting insulin dose to exercise). Self-esteem was assessed using 8-point scale 64% of patients with diabetes self-reported good adherence to their insulin regimen. High self-esteem was associated with self-reported adherence to insulin regimen (p=0.005)
Kampman et al. 200161 Prospective interview study 100 hospital inpatients and outpatients receiving neuroleptic medication (mean age 37.6 years, age range 16-63 years) Discrepancy in compliance assessments performed by physicians and patient self-reported compliance Patient characteristics explaining discrepancy in assessments included mania, high neuroleptic dose, female gender, low education level, and attitudes toward neuroleptic treatment
Jokisalo et al. 200146 Cross-sectional mail surveys and health examination of general practice patients 1782 (80%) hypertensive patients from 26 health centres during one week in 1996, of whom 1561 on antihypertensive medication 14 self-reported problem indices formulated from 82 items about experiences in relation to hypertension treatment Problems included lack of motivation for follow-up (72%), difficultly accepting being hypertensive (66%), careless attitude towards hypertension (63%), lack of information (56%), and hopelessness, impact on sexual function and lack of support by health care personnel (33%). The least frequent problems included modification of dosage instructions. The number of problems per person ranged from 0-13, mean 4.9
Enlund et al. 200143 Cross-sectional survey of patients of 9 community pharmacies 971 patients aged <75 years receiving long-term treatment with antihypertensive medications between May and September of 1996, of whom 482 (56%) completed the questionnaire Self-reported self-initiated modification of dosage instructions (often, sometimes, no) with often and sometimes combined to form a dichotomous variable (modifiers and non-modifiers). Self-reported hypertension treatment problems (symptoms of hypertension, adverse drug effects, perceived negative patient role, memory problems, interference with daily routines, had to give up pleasant activities/interference with hobbies) 31% of males and 21% of females self-reported they modified their dosage instructions. Patients with 3+ hypertension treatment problems and those with an academic education were significantly more likely to have modified their dosage instructions than those without problems (OR 4.79, 95% CI 2.36-9.73) and (OR 2.69, 95% CI 1.27-5.70)
Lahdenperä & Kyngäs. 200148 Interviews with participants in a compliance intervention comprising a multimedia computer program and lifestyle counselling 21 patients with hypertension aged 16-64 years who had been taking medication for less than one year or not at all Meaning of compliance to patients with hypertension, and attitude of patients toward their illness (S) Four levels of compliance detected (subconscious, cognitive, action and attitudinal). Four types of attitudes toward illness detected (careless, serious, well adjusted, frustrated)
Toljamo & Hentinen. 200166 Cross-sectional mail survey 213 people with insulin-treated diabetes aged 17-65 years who visited health centre or hospital (76% response rate) Adherence to self-care, difficulties in self-care, perceived receipt of social support (S) 10% of respondents reported sometimes, often or always experiencing difficulties with insulin treatment at home, while 43% experienced difficulties when not at home. Insulin therapy was not perceived as the most difficult aspect of self-care. There was an overall relationship between poor self-care and poor metabolic control
Toljamo & Hentinen. 200167 Cross-sectional mail survey 213 people with insulin-treated diabetes aged 17-65 years who visited health centre or hospital (76% response rate) Adherence to self-care according to prescribed health regimen (S) 84% of respondents accomplished daily insulin injections as scheduled, while a further 15% accomplished near daily insulin injections. Respondents were less adherent to urine testing, exercise and dietary recommendations
Kyngäs. 200162 Cross-sectional postal survey using health register as a sampling frame 1200 children and adolescents with epilepsy, juvenile rheumatoid arthritis, insulin-dependent diabetes mellitus and asthma Compliance (including medication compliance) coded as a dichotomous variable (compliant versus non-compliant) Support from nurses OR 7.28 (95% CI 3.95-13.42), energy and willpower OR 6.69 (95% CI 3.91-11.46) and good motivation OR 5.28 (95% CI 3.02-9.22) were the strongest predictors of compliance
Kyngäs. 200156 Cross-sectional postal survey using health register as a sampling frame 300 children and adolescents with epilepsy aged 13-17 years (77% response rate) Compliance (including medication compliance) coded as a dichotomous variable (compliant versus non-compliant) Support from physicians OR 10.56 (95% CI 2.06-15.22), support from parents OR 10.47 (95% CI 2.19-14.77) and motivation OR 9.77 (95% CI 2.47-13.86) were the strongest predictors of compliance
Kyngäs. 200057 Cross-sectional postal survey using health register as a sampling frame 300 children and adolescents with epilepsy aged 13-17 years (77% response rate) Compliance with health regimens (including medication compliance). Compliance divided into 3 categories; good, satisfactory and poor 37% categorised as having good medication compliance, and 31% poor medication compliance
Kyngäs. 200065 Cross-sectional postal survey using health register as a sampling frame 300 children and adolescents aged 13-17 years with diabetes mellitus (97% response rate) Compliance with health regimens (including medication compliance). Compliance divided into 3 categories; good, satisfactory and poor 81% of children or adolescents reported good compliance with insulin treatment. 48% of children or adolescents reported their treatment was planned together with physicians. 49% of children or adolescents reported their treatment was planned together with nurses
Kyngäs. 200069 Cross-sectional postal survey using health register as a sampling frame 300 children and adolescents aged 13-17 years with juvenile rheumatoid arthritis (91% response rate) Compliance (including medication compliance) divided into 3 categories; good, satisfactory and poor 3% categorised as having good medication compliance, 52% satisfactory medication compliance and 45% poor medication compliance

Abreviations: OR=odds ratio; CI=confidence interval; HR=hazard ratio; PANSS=positive and negative syndrome scale; DSM-IV =Diagnostic and Statistical Manual of Mental Disorders (Version 4); (S)=compliance, adherence or persistence was a secondary outcome measure only.

The terminology used in the table (compliance/adherence/persistence) corresponds to the terminology used by the authors of the above studies