Inhaler Technique in Low-Income, Inner-City Adults with Uncontrolled Asthma

Abstract

BACKGROUND:

Poor inhaler technique has been shown to be associated with less asthma control and increased healthcare utilization. Little is known about the impact of inhaler technique on the most vulnerable patients.

OBJECTIVE:

This study examined inhaler technique in low-income, inner-city adults with uncontrolled asthma.

METHODS:

Inhaler technique data and other patient characteristics were evaluated in adults drawn from two studies conducted at the University of Pennsylvania. Subjects were from low-income Philadelphia neighborhoods and had uncontrolled asthma. Baseline characteristics were collected. Inhaler technique was rated by research coordinators who were trained with written materials.

RESULTS:

In 584 adults, 56% of metered dose inhaler (MDI) users and 64% of dry powder inhaler (DPI) users had adequate visually-assessed inhaler technique. Inhaler technique did not vary by reading comprehension or numeracy levels.

CONCLUSION:

In this group of patients with uncontrolled asthma, visually-assessed inhaler technique was adequate in more than one-half. While incorrect inhaler technique is generally common and must be routinely addressed, this study suggests that other factors that lead to poor control must be identified.

INTRODUCTION

Prior research suggests that anywhere from one-half to all patients with asthma and COPD do not correctly use their inhalers, with recent pooled summaries finding that 87% and 61% of metered dose inhaler (MDI) and dry powder inhaler (DPI) users make at least one error, respectively.¹ Poor inhaler technique reduces drug delivery and bronchodilation^2,3,4,5,6 and has been associated with less disease control for MDI and DPI users. ^{7,8,9,10,11,12,13} Recently, Price and colleagues (2017) went further by linking specific errors to asthma outcomes in the CRITIKAL study.¹⁴ With MDIs, coordination of actuation and inhalation was associated with asthma control. With DPIs, inadequate inspiratory effort predicted worse control.

Among individuals with uncontrolled asthma, the relative importance of inhaler technique compared to other variables is unknown. Tobacco smoke exposure,¹⁵ minority and lower socioeconomic status,^16,17 inadequate insurance coverage,¹⁸ and poor-quality housing^19,20 are also important predictors of disease burden. Health literacy is now recognized as a predictor of asthma outcomes.^{21,22,23,24,25} Numeracy is a component of health literacy comprising practical quantitative skills, distinguished from comprehension of prose, necessary for understanding concepts like dosing, timing of medications, and other quantitative aspects of self-managing a chronic disease. Numeracy has been associated with Emergency Department visits and hospitalizations for asthma.²⁶

The purpose of our analysis was to assess inhaler technique in high-risk adults with uncontrolled asthma, as inhaler technique is one of the modifiable risk factors for poor disease control. We predicted that these patients would demonstrate high overall rates of inhaler technique errors. We also hypothesized that patients with lower health literacy would have poorer technique. Finally, we considered technique in patients with both MDI and DPI inhalers, whom we called dual users. We assessed inhaler technique visually as in a routine medical evaluation.

METHODS

586 de-identified data were drawn from two studies performed at the University of Pennsylvania and approved by the institutional review boards of the University of Pennsylvania and the Corporal Michael J. Crescenz Veterans Administration Medical Center. Participants were adults from low-income neighborhoods in Philadelphia. 301 adults had uncontrolled asthma, defined as requiring prednisone, an Emergency Department visit, or a hospitalization for asthma in the past 12 months. An additional 285 adults had moderate to severe asthma, an FEV₁ < 80% predicted (at or within 3 years of enrollment), and evidence of reversible airway obstruction (improvement with bronchodilator defined as either: an increase of ≥15% and 200 ml in FEV₁ with asthma treatment over 3 years; or an increase of FEV₁ or FVC ≥12% and 200 ml within 30 minutes of 4 puffs of albuterol by MDI or 2.5 mg albuterol by nebulizer). All adults in the second group had uncontrolled asthma by FEV₁ percent predicted, and most by Asthma Control Questionnaire (ACQ), according to National Asthma Education and Prevention Program Expert Panel Report (EPR) 3 guidelines.^27,28 Therefore, we considered all 586 adults to have uncontrolled asthma.

All were prescribed an inhaled corticosteroid and technique with that inhaler was rated at enrollment by research coordinators. Baseline patient characteristics were collected. MDI technique was assessed using albuterol; all DPI subjects used Advair Diskus® inhalers. MDI technique rating was derived from inhaler guidelines from the EPR 3,²⁷ and DPI technique from Advair Diskus® package insert instructions, using common steps across each device type. MDI technique had 7 steps; DPI technique had 6 steps (Table I). Each step was rated ‘yes’ (coded as 1) or ‘no’ (coded as 0). After scoring, incorrect steps were corrected to benefit the users. Research coordinators scored technique visually. The research coordinators for one group were recent college graduates interested in a healthcare career. The research coordinators for the other group were community health workers who were high school graduates with three years’ work experience. The coordinators were trained in obtaining spirometry by the clinicians and project manager, and also using written materials, including the scoring system presented in Table I. They were observed in role-playing and then with patients. Their evaluations were checked by the project manager periodically.

Table I.

Scoring systems for Metered Dose Inhaler (MDI) and Dry Powder Inhaler (DPI) techniques among 584 patients with uncontrolled asthma.^*

MDI technique	DPI technique
Canister shaken	Exposes mouthpiece
Exhales before actuating inhaler	Cocks the trigger (activates the Diskus®)
Actuates inhaler at the start or within 1 second of the start of inhalation	Holds Diskus® horizontally while inhaling
Actuates only once per inhalation	Inhales deeply
Inhalation takes place over 3-5 seconds	Holds breath for 6 seconds
Position of inhaler appears adequate	Does not blow into Diskus®
Patient holds breath for 6-10 seconds

^*Each step was rated ‘yes’ (coded as 1) or ‘no’ (coded as 0). Incorrect steps were corrected. Thus MDI score ranged from 0-7 and DPI score from 0-6.

Inhaler technique was scored from 0 to 100 as a percentage of possible correct steps. Owing to the very skewed nature of the data on inhaler technique, we grouped the data into four ordered groups. The four categories reflected simple thresholds to achieve adequate numbers of patients per category. 1 = 0% to less than 50%; 2 = 50% to less than 75%; 3 = 75% to less than 100%; 4 = 100%. Reading comprehension was assessed using the Short Test of Functional Health Literacy in Adults (S-TOFHLA) and was treated as a binary variable (greater than or equal to 23 versus less than 23).²⁵ Numeracy was assessed using the Asthma Numeracy Questionnaire (ANQ) and was also dichotomized (greater than 2 versus less than or equal to 2).²⁵ These decisions reflected common thresholds for scoring.^29,30 We used the Cochran-Mantel-Haenszel test to assess whether the scores of inhaler technique for both MDIs and DPIs differed by level of reading comprehension or numeracy.

We also examined the possibly critical elements of each inhaler type. For MDIs we included “exhales before actuating inhaler,” “actuates inhaler at the start or within 1 second of the start of inhalation,” “actuates only once per inhalation,” and “inhalation takes place over 3-5 seconds,” and for DPIs, “inhales deeply.”¹⁴

RESULTS

584 of 586 subjects had complete inhaler technique data. Of these, 579 had S-TOFHLA data and all had numeracy data. Baseline patient characteristics are detailed in Table II. Of the 584 evaluable subjects, 459 were rated using MDIs; 192 were rated using DPIs (Table III). 67 were dual users (Table IV). Overall, 348 subjects (60%) had adequate visually-assessed inhaler technique. Among MDI-only users, 172 (44%) made at least one error. Commonly failed steps included exhaling before actuating the inhaler (34%), actuating only once per inhalation (13%), and breath hold for 6-10 seconds (19%). Among DPI-only users, 45 (36%) made at least one error. The only commonly failed step was breath hold for 6 seconds (33%). There was no relationship between inhaler technique and asthma control based on the Asthma Control Questionnaire.

Table II.

Characteristics including ethnicity, literacy, and comorbidities of 586 adults with uncontrolled asthma.

Characteristic	n (%)
Demographics
Age (min 18, max 93 years)	50 +/− 14
Female	461 (79%)
Race
Black/African American	412 (70%)
White	96 (16%)
Ethnicity: Hispanic/Latino	89 (15%)
Educational attainment
8th grade or less	24 (4%)
Some high school	102 (17%)
High school grad or G.E.D.	194 (33%)
Some college or trade school	141 (24%)
College graduate	125 (21%)
Disease severity
Mean FEV1 % predicted (min 11, max 118)	69 +/− 20
Mean ACQ* (min 0.0, max 5.7)	2.4 +/− 1.2
ACQ ≤ 1.5 (adequate)	151 (26%)
Hospitalized for asthma past 1 year	224 (38%)
ER for asthma past 1 year	366 (62%)
Health Insurance
Medicaid	287 (49%)
Healthcare literacy
Mean S-TOFHLA† (min 0, max 38)	27 +/− 12
S-TOFHLA > 23 (adequate)	432 (74%)
Mean ANQ‡ (min 0.0, max 4.0)	1.7 +/− 1.4
ANQ 3-4 (adequate)	175 (30%)
Comorbidities
BMI
<18.5 (underweight)	4 (0.7%)
18.5-24.9 (normal weight)	73 (12%)
25-29.9 (overweight)	126 (22%)
>30 (obese)	382 (65%)
Hypertension	325 (55%)
Diabetes	151 (26%)
Ever smoker	307 (52%)
Current smoker	132 (23%)

^*Asthma Control Questionnaire (ACQ) range = 0-6 (a 6-item questionnaire, with lower score indicating better control; inadequate control is a score >1.5).^44,45,46

^†Short Test of Functional Health Literacy in Adults (S-TOFHLA) is a reading comprehension test with range from 0-36, (score ≥ 23 is adequate and score < 23 is inadequate).

^‡Asthma Numeracy Questionnaire (ANQ) is a 4-item questionnaire⁴ with range from 0-4 (0-2 is low; 3-4 is adequate).

Table III.

Number and percent of patients accomplishing each element of Metered Dose Inhaler (MDI) technique (n = 459) and Dry Powder Inhaler (DPI) technique (n = 192).^*†

Element of MDI technique (n = 459)	n (%)
Canister shaken	339 (86%)
Exhales before actuating inhaler	260 (66%)
Actuates inhaler at the start or within 1 second of the start of inhalation	368 (94%)
Actuates only once per inhalation	342 (87%)
Inhalation takes place over 3-5 seconds	354 (90%)
Position of inhaler appears adequate	374 (95%)
Patient holds breath for 6-10 seconds	317 (81%)
Element of DPI technique (n = 192)	n (%)
Exposes mouthpiece	123 (98%)
Cocks the trigger (activates the Diskus®)	120 (96%)
Holds Diskus® horizontally while inhaling	122 (98%)
Inhales deeply	123 (98%)
Holds breath for 6 seconds	85 (68%)
Does not blow into Diskus®	120 (96%)

^*2 subjects had missing inhaler technique data.

^†67 of 584 subjects had both MDI and DPI data.

Table IV.

Technique for dual MDI/DPI users (n = 67)

Element of MDI technique	n (%)
Canister shaken	63 (94%)
Exhales before actuating inhaler	55 (82%)
Actuates inhaler at the start or within 1 second of the start of inhalation	59 (88%)
Actuates only once per inhalation	55 (82%)
Inhalation takes place over 3-5 seconds	57 (85%)
Position of inhaler appears adequate	62 (93%)
Patient holds breath for 6-10 seconds	56 (84%)
Element of DPI technique	n (%)
Exposes mouthpiece	65 (97%)
Cocks the trigger (activates the Diskus®)	65 (97%)
Holds Diskus® horizontally while inhaling	62 (93%)
Inhales deeply	60 (90%)
Holds breath for 6 seconds	59 (88%)
Does not blow into Diskus®	62 (93%)

Statistical Analysis Software (SAS) was used for analysis.³¹ Of 581 asthmatic adults who took the S-TOFHLA, 432 (74%) had a score greater than or equal to 23, indicating adequate reading comprehension. About sixty percent of all patients achieved visually-assessed adequate inhaler technique. Of the 586 patients who took the Asthma Numeracy Questionnaire (ANQ), only 175 (30%) had a score greater than 2 for the four items. In both MDI and DPI analyses, the fraction of patients who had the highest inhaler scores was only slightly greater in the higher reading comprehension and numeracy groups. Thus, there were no differences in the distribution of inhaler technique quartile scores by either reading comprehension or numeracy levels (Tables V–VI). The tables show that the distributions are very similar by numeracy and reading comprehension levels. There was also no association between critical elements of MDI or DPI technique and either reading comprehension or numeracy levels.

Table V.

The lack of association of inhaler technique with binary reading comprehension (p = 0.77).^§

Reading Comprehension	Elements of inhaler technique†
n* Row percent	<50% of elements correct	50-74% of elements correct	75-99% of elements correct	100% of elements correct	Total
S-TOFHLA‡ score <23	10	25	26	86	147
	6.8	17.0	17.7	58.5
S-TOFHLA score ≥23	29	67	79	257	432
	6.7	15.5	18.3	59.5
Total	39	92	105	343	579

^*7 patients out of 586 (1.2%) were missing inhaler technique data (2 missing) and literacy data (5 missing).

^†Column categories for inhaler technique were created a priori to correspond to understandable categories of % correct and to produce adequate minimum cell counts.

^‡Short Test of Functional Health Literacy in Adults (S-TOFHLA). Scores lower than 23 reflect marginal or worse literacy and inability to comprehend information in health educational pamphlets.

^§Cochran-Mantel-Haenszel test of differences in row mean scores.

Table VI.

The lack of association of inhaler technique with binary numeracy (p = 0.47). ^§

Numeracy	Elements of Inhaler Technique†
n* Row percent%	<50% of elements correct	50-74% of elements correct	75-99% of elements correct	100% of elements correct	Total
ANQ‡ score ≤ 2	28	70	68	243	409
	6.9	17.1	16.6	59.4
ANQ score ≤ 2	11	22	37	105	175
	6.3	12.6	21.1	60.0
Total	39	92	105	348	584

^*2 patients out of 586 (0.3%) were missing inhaler technique data.

^†Column categories for inhaler technique were created a priori to correspond to understandable categories of % correct and to produce adequate minimum cell counts.

^‡Asthma Numeracy Questionnaire (ANQ). Scores of 2 or less reflect inadequate quantitative self-management skills.

^§Cochran-Mantel-Haenszel test of differences in row mean scores.

DISCUSSION

Despite having uncontrolled asthma, the patients in this analysis demonstrated better inhaler technique than was expected and when compared to the asthma population as a whole. Specifically, the percentage of patients committing at least one error was only 44% for MDIs and 36% for DPIs, compared to 87% and 61% reported by Chrystyn et al (2017).¹ We can postulate why technique was better than predicted. Perhaps these subjects, drawn mostly from university clinics, received adequate teaching of technique. Or maybe their high rate of Emergency Department and hospital visits resulted in reinforcement of proper technique: In the uncontrolled asthma group, patients by definition had interfaced with the healthcare system in the previous year; in the moderate-to-severe asthma group, almost half had at least one asthma-related Emergency Department visit in the past year, with one-third being hospitalized for asthma. It is well-established that provider education and demonstration of inhaler use improves technique,^32,33 affirming guidelines that recommend providers review technique at each clinic visit.^27,34 Importantly, old age³⁵ and low health literacy³⁶ are not barriers to learning and retaining proper technique. Our results further the proposition that vulnerable patients can demonstrate good inhaler technique.

The seeming dissociation between subjects’ asthma control and inhaler technique is counterintuitive, and may be explained by important baseline characteristics in our patients. 23% of patients were current smokers. The mean BMI was 35.1, and obesity is known to be associated with more severe asthma.³⁷ Suboptimal living conditions, ubiquitous in poor Philadelphia neighborhoods, is possibly a critical factor. Lastly, 49% of patient were on Medicaid, far above the national average,³⁸ and individuals covered by Medicaid have higher rates of poor asthma control.³⁹

Our results must be interpreted with caution. Chrystyn and colleagues (2017) found a high degree of between-study heterogeneity in their meta-analysis, likely owing to a wide variety of inhaler types and checklists used, varying study types, and wide variability in patient characteristics and assessor skill levels.¹ Additionally, most studies of adults recruited both asthma and COPD patients and did not validate these diagnoses nor separate subjects by disease type. A strength of our analysis was the inclusion of patients with uncontrolled asthma, requiring a physician’s diagnosis of asthma and a hospitalization, Emergency Department visit, or prednisone use in the past year. The subset with moderate-to-severe disease required both a physician’s diagnosis and an FEV₁ less than 80% predicted combined with improvement with a bronchodilator.

Another uncertainty is defining proper technique. Checklists are not standardized and furthermore may not be applicable to all inhalers, as novel devices are continually added to the market. For a given device and checklist, investigators’ threshold for defining good or bad technique often varies. A review on technique by Mahon and colleagues (2017) did not assess absolute error rates but instead included studies that noted “failure/misuse” rates, the definition of which differed across studies.⁴⁰ Good technique may alternatively be defined by the rate of critical errors, or those errors associated with worse disease control. A growing body of research is identifying critical errors in asthma and COPD,^8,13,14 which may ultimately become the basis for grading technique.

Our study has several limitations. We used commonly accepted checklists and instructions to assess technique, but we did not consider all possible errors. The CRITIKAL study assessed fourteen errors for each inhaler type, including an evaluation of head tilt and breathing out to empty lungs before inhalation for both inhaler types; and for DPIs, downward holding of the device after cocking the trigger. Errors omitted from our checklists were among the most common in the CRITIKAL study¹⁴ and other studies.^13,41,42,43 Therefore, our findings may have been biased by omitting potentially important errors. In part, this limitation reflects a broader issue of no universally accepted “correct” technique, and there is wide inter-study variability in inhaler checklists.¹ It is noteworthy the CRITIKAL study was published in 2017, well after our studies began. Additionally, our study was not a technical study. We did not measure airflow with a validated device. Instead, we observed patients using inhalers as they would be observed by a health professional. Finally, we were not able to determine if inhaler technique differed in patients when not under observation (Hawthorne effect).

Our study population is exceptional, including adults with high morbidity. While we found that there is uncertainty in measuring technique for all the reasons discussed, good technique predominated among the patients in our study. While incorrect technique is common and must be addressed as part of routine asthma care, other factors also contribute to poor asthma outcomes and must be explored.

What is already known about this topic?

Inhaler misuse is common and can be associated with uncontrolled asthma. In low-income inner-city adults, a group with high asthma morbidity, the rate of inhaler misuse is unknown.

What does this article add to our knowledge?

In this cohort of patients with uncontrolled asthma, inhaler technique was better than expected.

How does this study impact current management guidelines?

This study illustrates that while correct inhaler technique is necessary, incorrect technique cannot explain the poor disease control in our patient population. In individuals with poorly controlled asthma, other factors contributing to disease morbidity must be considered.

Abbreviations used

ACQ

Asthma Control Questionnaire

ANQ

Asthma Numeracy Questionnaire

BMI

body mass index

COPD

chronic obstructive pulmonary disease

DPI

dry powder inhaler

EPR

Expert Panel Report

FEV₁

forced expiratory volume in one second

FVC

forced vital capacity

MDI

metered dose inhaler

S-TOFHLA

Short Test of Functional Health Literacy in Adults