TABLE 2

Definition and measurement of adherence behaviours in the studies reviewed

First author [ref.]	Year	Adherence definition/term	Assessment method	Details	Validity/reliability
Tettersell [32]	1993	Taking inhalers as prescribed	Patient-reported, single item	One item: “do you take your inhalers as prescribed?”; four response options: “always”, “majority of the time (8 out of 10 doses)”, “about half of the time”, “only during or following an attack”	NR
Bosley [33]	1995	Noncompliance; taking <70% of prescribed doses or omitting all doses for ≥1 week	Electronic monitoring	Turbuhaler Inhalation Computer; computed for two 6-week periods as (no. of doses taken)/(no. of doses prescribed)×100	NR
Apter [34]	1998	Use of ICS in the last 35 days	Electronic monitoring	MDI_log, last 35 of 42 days considered, computed for 12-h periods as (recorded − prescribed actuations)×100; mean truncated adherence computed per subject; dichotomised (<or>70%)	NR
Bennett [35]	1998	Adherence to preventive ICS use	Patient-reported, published scale	RAAS [83]	α=0.75
Chambers [36]	1999	Frequency of ICS use	Patient-reported, single item	Item content not specified, four response options: “I use it at least twice a day almost every day”, “some days I use it at least twice, but on other days I don't use it at all”, I used to use it, but now I don't”, “I never used it”; dichotomised into “regular, twice daily” and “less than regular”	NR
Schmaling [37]	2000	As-needed medication use	Canister weighting	Total number of medication inhalations for each day in the prescription period	NR
		Daily medication adherence	Canister weighting	Predicted use (no. days × no. puff per day) compared to actual use; computed as percent of prescribed medication used	NR
Horne [38]	2002	Medication adherence	Patient-reported, published scale	MARS [38]	α=0.85
van Schayck [39]	2002	Medication compliance rate	Canister weighting	Medication used as a percentage of medication prescribed	NR
Apter [40]	2003	Use of ICS in the last 42 days	Electronic monitoring	MDI_log, 42 days, computed for 12-h periods as (recorded − prescribed actuations)×100; mean truncated adherence computed per subject; divided into four categories (<20%, 20–<50%, 50–<75%, 75–100%)	NR
Jessop [41]	2003	Adherence to preventative inhaled medication in the last 3 months	Patient-reported, published scale (adapted)	RAAS [83] and two extra items on accidental nonadherence	α=0.92
Labrecque [42]	2003	Conformity of SABA prescription use with accepted good use criteria	Medical (refill) records	Dichotomous, good use criteria: for SABA with no ICS use, the interval between the targeted SABA prescription and the following refill corresponds to a maximum daily use of two inhalations; for SABA with ICS use, the criterion above, and a daily ICS dose below a fixed threshold	NR
Nishiyama [43]	2003	Reliever compliance	Patient-reported, interview	Patients were required to state the drugs and dosage regimens they used; their reports were compared with prescription information; three values were coded: “good”; “overused”; “underused” (first two also applied to reliever)	NR
		Preventer compliance	Patient-reported, interview		NR
Balkrishnan [44]	2005	Adherence to controller pharmacotherapy	Medical (refill) records	Computed as: (days of prescription supply dispensed)/(days between prescription refills − number of days person was hospitalised); dichotomised as compliant (0.5–1.5) or not	NR
Lacasse [45]	2005	Non-compliance	Electronic monitoring	MDI_log; calculated for 12 weeks daily as proportion of prescribed daily dose actually inhaled; dichotomised as compliant (>75%) or not	NR
Stempel [46]	2005	Asthma medication refill rate	Medical (refill) records	Number of 1-month supply during the 12-month post-index period	NR
		Number of treatment days	Medical (refill) records	For monotherapy: total days supplied of medication For combination: total days supplied of ICS	NR
		SABA refill rates	Medical (refill) records	Number of 1-month supply during the 12-month post-index period	NR
Bender [47]	2006	Adherence to ICS/LABA	Medical (refill) records	Total days supplied during follow-up period	NR
		Persistence	Medical (refill) records	Time to discontinuation computed as number of days from index date to date preceding the pre-specified gap when supply was exhausted	NR
Chatkin [48]	2006	Compliance	Canister weighting	(Total quantity of medication used)/(quantity prescribed, i.e. three canisters in 3 months); dichotomised as compliant (>85%) or not	NR
Hasegawa [49]	2006	Drug compliance	Medical (prescription and refill) records	Computed for 6 months as (number of medicines dispensed)/(number of medicines prescribed)×100; capped at 100%	NR
Marceau [50]	2006	Persistence versus discontinuation: having prescriptions continuously renewed within the period	Medical (refill) records	Computed as the sum of three times the duration of the current prescription (in days) plus all overlaps accumulated since therapy start; discontinuation date was the end date of the last filled prescription plus all overlaps	NR
Ohm [51]	2006	Use of ICS+LABA	Electronic monitoring	Advair diskhaler; computed as (number of counted doses)/(number of prescribed doses)×100; dichotomised as good adherence (≥80%) or not	NR
		Medication adherence	Patient-reported, published scale	MARS [38]	NR
Tavasoli [52]	2006	Compliance to prescribed MDI drugs	Patient-reported, interview	Four items: "do you use your prescribed spray (MDI drug) regularly?", "have you ever had any history of not using your spray?", "do you still use your last prescribed spray?", "how do you use your spray? Show me"; response scales from 0 to 4	NR
Ulrik [53]	2006	Intentional nonadherence	Patient-reported, single item	One item: “how often do you decide not to take your controller medication?”; five response options: “almost every day”, “a couple of times every week”, “a couple of times every month”, “a couple of times every year”, “hardly ever”	NR
		Adherence	Patient-reported, single item	Item not specified; responses reported on a three-level scale: taking controller therapy as prescribed, less, or more than prescribed	NR
Williams [54]	2007	ICS adherence	Medical (refill) records	(Cumulative days supplied)/ (total number of days between refills for 1-year study period); analyses performed also with adherence stratified (0%, 0%–80%, ≥80%)	NR
Williams [55]	2007	Primary non-adherence	Medical (refill) records	No prescription fill information recorded for 3 months after index prescription	NR
		ICS adherence	Medical (refill) records	Computed as (total days supplied)/(number of days of observation)×100; adherence stratified (0%, 0–80%, ≥80%)	NR
Breekveldt-Postma [56]	2008	Persistence during the first year	Medical (refill) records	Computed as number of days from start to time of first failure to continue renewal of initial prescription, based on (number of units dispensed)/(number of units to be used per day as defined in pharmacy)	NR
Janson [57]	2008	ICS nonadherence during the last 14 days	Patient-report, interview	Nursing home assessment of ICS prescription and use, based on inspection of current asthma medication and two questions: “How many puffs and how many times per day did your doctor tell you to use this?”, “During the past 14 days, how many puffs and how many times per day have you used this?”; dichotomised as adherent (≥7 days of use in previous 14 days) or not	NR
		SABA or LABA overuse	Patient-reported, interview	Nursing home assessment on SABA and LABA prescription and use, dichotomised as overuse (average >8 puffs of SABA or >2 puffs of LABA -single or combination- per day) or adherent	NR
Martínez-Moragón [58]	2008	Frequency of ICS use	Patient-reported, single item	One item, not specified, adapted after (37); four response options, from “never” to “at least twice a day almost every day”, dichotomised into “almost every day” versus “rarely if ever”	NR
McGann [59]	2008	“How closely an individual's medication taking behaviours, as measured by the DOSER, approximated prescribed use instructions provided by the healthcare provider”	Electronic monitoring	DOSER; ratio of the number of observed correct prescribed use days between day 3 and 14	Agreement with other measures (not specified) 84.32%
Menckeberg [60]	2008	Medication acquisition	Medical (refill) records	(Total days supplied)/(total number of days from first and last refill date)×100
		Medication adherence	Patient-reported, published scale	MARS [38]	α=0.81
Wells [61]	2008	ICS adherence; the proportion of time that the patient had medication available during last 6 months	Medical (refill) records	(Total days supplied)/(number of days of observation)×100	NR
Axelsson [62]	2009	Medication adherence	Patient-reported, published scale	MARS [38]	α=0.71
Bae [63]	2009	Prescription refill adherence	Medical (refill) records	(Number of ICS refills)/12×100; categorised as appropriate use (>80%), underuse (50–80%), or extreme underuse (<50%)	NR
		Subjective self-reported adherence	Patient-reported, single item	One item: “how often did you take your ICS as prescribed for last 1 year?”; response on a visual analogue scale from 0% to 100%; categorised as appropriate use (>80%), underuse (50–80%), and extreme underuse (<50%)	NR
Laforest [64]	2009	Intentional interruption	Patient-reported, single item	Six items included: 1) accidental interruption, 2) intentional interruption when feeling better, 3) intentional interruption when feeling worse, 4) reduced use when feeling better, 5) more frequent use of ICS in case of preliminary signs of asthma attack, and 6) intentional changes of doses independently of physician; analyses performed on intentional (when feeling better) and accidental interruption	NR
		Accidental interruption	Patient-reported, single item
Ponieman [65]	2009	Medication adherence	Patient-reported, published scale	MARS [38]; dichotomised as good adherence (≥4.5) or not	α=0.86
Friedman [66]	2010	Prescription fills	Medical (refill) records	Total number of prescription refills during the post-index period	NR
		Percentage of days covered	Medical (refill) records	(Number of days patients had medication on hand)/(total number of post-index days=365)×100	NR
Takemura [67]	2010	Self-reported adherence to inhalation regimen	Patient-reported, published scale (adapted)	Modification of RAAS [83] concerning the use of inhaled controller medications; mean adherence score computed; dichotomised as good adherence (≥4.0) or not	NR
Bolman [68]	2011	Medication adherence	Patient-reported, published scale	MARS [38]	α=0.89
Emilsson [69]	2011	Medication adherence	Patient-reported, published scale	MARS [38]	α=0.77
Small [70]	2011	Physician-perceived compliance; “the extent to which the patients are perceived to follow their physician's prescribing instructions and advice”	Physician-reported, bespoke scale	Two items (not specified) on physician-perceived patients' compliance regarding frequency of use and inhaler use; five response options from “not at all compliant” to “fully compliant”	α=0.92
Suzuki [71]	2011	ICS adherence	Medical (prescription and refill) records	Ratio of doses dispensed in the pharmacy divided by prescribed doses documented in medical charts	NR
Foster [72]	2012	Adherence with ICS/LABA	Electronic monitoring	Smart inhaler; daily adherence calculated as (no. recorded actuations/no puffs prescribed)×100, capped at 100% and averaged for the last 4 weeks of 2 months monitored	NR
			Patient-reported, published scale	Morisky adherence scale [84]	NR
			Patient-report, single item	Estimation of own inhaler use (days/week and puffs per day) in the last 4 weeks	NR
Ahmedani [73]	2013	ICS adherence	Medical (prescription and refill) records	(Total days supplied)/(3-month observation period)×100	NR
Axelsson [74]	2013	Medication adherence	Patient-reported, published scale	MARS [38]	α=0.75
Price [75]	2013	ICS adherence	Patient-reported, published scale	MARS [38], categorised as “low”(“often” or “always” response to any question), “borderline” (´sometimeś responses to > 1 question), and “good” (any other answer)	NR
Price [76]	2013	ICS adherence	Medical (prescription) records	(Total days supplied)/(365-day observation period)×100	NR
Schatz [77]	2013	Questionnaire low adherence	Patient-reported, published scale	Response to “how often are you actually taking your ICS medication now” compared to response to “based on your doctor's most recent instructions, how often were you advised to be taking your ICS medication now” (less frequently)	NR
		Percent of days covered	Medical (refill) records	Days’ supply of dispensed canisters over the follow-up at 3, 6, and 12 months	NR
Wells [78]	2013	ICS adherence	Medical (prescription and refill) records	Continuous multiple-interval measure of medication availability equals number of days' supply for each fill/total number of days between the present and next fill; averaged for the observation period	NR
Baddar [79]	2014	Compliance with controller treatment	Interview cross-checked with electronic patient records	Good equals taking 100% of daily prescribed medication and ≤2 missed doses/administrations per week; partial equals taking more or less than their daily prescribed medication; poor equals any other inhaler use patterns	NR
Federman [80]	2014	ICS adherence	Dose count	Review of dose counters for all dry powder inhaler devices during the first 3 months and 30 days after each new prescription; dichotomised as <80% and ≥80%	NR
Taylor [81]	2014	Adherence to ICS prescriptions	Medical (prescription) records	Prescription possession ratio: (number of days prescribed during calendar year)/(number of days in the interval)×100	NR
van Steenis [82]	2014	ICS adherence	Patient-reported, published scale (adapted)	Morisky adherence scale [84], adapted	NR
		ICS adherence	Medical (refill) records	Proportion of days covered: (number of days' supply)/(365 or truncated if medication gap ≥182)×100; dichotomised as <80% and ≥80%	NR

NR: not reported; ICS: inhaled corticosteroids; RAAS: Revised Asthma Adherence Scale; MARS: Medication Adherence Rating Scale; MDI: metered-dose inhaler; SABA: short-acting β₂-agonists; LABA: long-acting β₂-agonists; α: Cronbach's α test.