Accuracy of Self-Reporting
Accuracy of self-reporting was measured by comparing mean self-reported sets per day to mean NC sets per day for both the feedback and control groups. Self-reported adherence (fewer than one, one, two, or three sets per day) was compared with the corresponding mean NC adherence category (fewer than 0.5, 0.5 to 1.49, 1.5 to 2.49, and greater than 2.5 sets per day).
The self-reporting of 44 percent of the feedback participants matched their NC adherence category, compared with only 25 percent in the control group (p<0.007). Control group participants also were more likely to overestimate their actual usage by at least one NC adherence category (71 percent) compared with the feedback group (49 percent) (p<0.002).
Canister Weighing vs the Nebulizer Chronolog
Eighty-one percent of both the feedback and control groups brought all four inhaler canisters back to the clinic at the follow-up visit. Table 3 shows the mean canister weight sets per day and mean NC sets per day for these participants. Canister weight change alone was not significantly different between the feedback group (2.33±0.67 sets per day) and the control group (2.32 ±0.99 [p = 0.92]), when canister weight change is defined in terms of sets of any two inhaler actuations irrespective of the timing of these actuations. The NC showed, however, that the feedback group averaged significantly more sets per day (2.07 ±0.60) than the control group (1.62 ±0.83 [p = 0.0003]), when sets were defined as the first two actuations within a defined time period (3 h). For both groups, mean sets per day determined by the NC were significantly lower than the mean sets calculated from the canister weights (p<0.0001).

Table 3—Comparison of Mean Sets per Day as Measured by Canister Weight and the Nebulizer Chronolog