Long-term Adherence to Topical Psoriasis Treatment Can Be Abysmal

A 1-year Randomized Intervention Study Using Objective Electronic Adherence Monitoring

H. Alinia; S. Moradi Tuchayi; J.A. Smith; I.M. Richardson; N. Bahrami; S.C. Jaros; L.F. Sandoval; M.E. Farhangian; K.L. Anderson; K.E. Huang; S.R. Feldman


The British Journal of Dermatology. 2017;176(3):759-764. 

In This Article


We enrolled 40 subjects with mild-to-moderate psoriasis (Table 1). Twenty of the 40 subjects completed the entire 12 months of the study (Fig. 1). The other half withdrew early due to worsening psoriasis (need for change of treatment) (n = 3), starting a concomitant medication (n = 3), not liking the medication (feels greasy, liking a stronger treatment, hard to use) (n = 4), burning side-effects (n = 1), loss to follow-up (n = 8) and fungal infection (n = 1). One of the caps did not record any data.

Adherence in the standard-of-care control group was low. In the first month, no medication was used on 37·4% of the days; over the last month of treatment, no medication was used on 50·9% of days. Drug holidays of 7 days or more without using the treatment were common, occurring in 35·2% of subjects in the first month and 42·8% in the 12th month of the study. The low levels of adherence were not due to clearing of the disease, as missed dosing was common in patients who still had psoriasis.

The intervention improved adherence. Overall, 35% of the prescribed number of doses were taken by the control group and 50% for the intervention group (P = 0·08). The percentage of days with correct dosing improved from 20% in the control group to 34% in the intervention group (P = 0·008, Fig. 2). Patients in the internet survey group had higher adherence (P = 0·03) in first month of the study and had numerically higher adherence until month 10 of the study. However, adherence in both standard-of-care and intervention groups declined over time (Fig. 3).

Figure 2.

Adherence rates in the control and intervention groups. A wide range of adherence behaviour was observed in both groups. The percentage of the days with doses taken (left) and the percentage of the correct doses taken (right) was greater in the intervention than in the control group (P = 0·01 and P = 0·008, respectively).

Figure 3.

Mean adherence in control and intervention groups over a year. Percentage of patients indicates the percentage of the number of the patients with the cross-section follow-up to the number of the whole cohort. Y: received internet survey, N: control group.

The between-group manova analysis was used to compare the PASI scores at baseline, and there was no significant difference between the control and intervention groups. There was a greater improvement in PASI in the intervention than in the control group at month 1 (1·61 vs. –0·12, P = 0·003), month 3 (2·50 vs. 0·79, P = 0·025) and month 12 (3·32 vs. 0·34, P = 0·038) (Fig. 4). We conducted the F-test manova between the two groups during all 12 months (longitudinally) using alcohol and smoking as covariates; the change in PASI over the 12-month study period trended towards statistical significance between the internet survey group and control group (P = 0·07, F-ratio = 2·89). There was little change in the IGA over the course of treatment, and there was no significant difference in IGA improvement between the control and intervention groups.

Figure 4.

Changes in mean Psoriasis Area and Severity Index scores. Greater improvement was seen in the intervention group than in the control group from baseline to the first month, third month and 12th month. The horizontal axis describes two groups.