Five interacting dimensions affecting adherence

A. Social and economic factors

• Poverty, illiteracy, lack of social support, poor access to services, high cost of transport, unstable living conditions, and family dysfunction associated with poor adherence
• Cultural beliefs
• War (adverse material and psychological effects)
• Degree of supervision of children and elderly patients

B. Health care team/system factors

• Quality of patient-provider relationship
• Cost of treatment
• Reliability of medication distribution systems
• Level of training and workload of healthcare providers
• Capacity for education and follow-up of patients
• Monitoring of performance of system
• Ability to establish community support and self-management capacity

C. Condition-related factors

• Severity of symptoms
• Level of disability
• Rate of progression of disease
• Availability of effective treatments

D. Therapy-related factors

• Complexity of the medical regimen (esp. dose frequency)
• Side-effects
• Duration of treatment
• Previous treatment failures
• Availability of medical support

E. Patient-related factors

• Knowledge and beliefs about their illness
• Risk perception
• Information and skills for self-management
• Motivation and self-efficacy
• Co-morbidities (esp. depression, alcohol and drug abuse

Trial

Intervention

Clinical outcomes

Adherence outcomes

Comments

FDC vs free combination

Tuberculosis

Su 2002, Taiwan¹¹

2 months Ritafer FDC, with Ethambutol + 4 months Rifinah FDC, with Ethambutol (n=57)

Sputum conversion At 2 months 95.0% vs 88.9% (p>0.05)

Compliance (not lost to follow-up or changed treatment) at 6 months 70.2 % vs 66.7% (p>0.05)

No difference demonstrated between the groups. Large loss to follow up (50% by 2 years) means that outcomes only measured in the selected group remaining in the study

RCT

vs free combination (n=48)

At 6 months 100% vs 100%

Radiological improvement At 2 years 92.3% vs 84.0% (p>0.05)

Geiter 1987, US¹²

2 months Rifater FDC + 4 months Rifamate FDC (n=169)

Sputum conversion At 8 weeks 86.6 vs 77.7%

Urine testing Pill counting Self-report At 8 weeks 96.5% vs 98.1% fully compliant (p>0.05)
At 6 months 88.5 vs 87.3 % fully compliant (p>0.05)

Early benefit in sputum conversion but no long-term difference in compliance.

RCT

vs free combination (n=532)

Absolute difference 8.9% (95% CI 1.1 to 16.7) (p<0.05)

Singapore Tuberculosis Service 1991,1999,^{25 26}

Ritafer FDC } streptomycin (one or two months) followed by isoniazid and rifampicin for remainder of 6 months (n=155)

Acceptability assessed by spontaneous complaints: same in both groups

Directly observed therapy High attendance in all groups

No demonstrated benefit of FDC

RCT

vs free combination of same drugs (n=155)

Adverse effects: similar in FDC and free groups

Slightly higher relapse rate at 2 years and 5 yrs in FDC groups

Culture negativity At 1 month 74 vs 72%, 76 vs 70%, 68 vs 62% At 2 months 100 vs 96%, 93 vs 91%, 95 vs 98%

Relapse during 18 months from end of treatment 6% vs 1% (p=0.04) Relapse at 5 years 7.9% vs 2.2% (p=0.03)

FDC vs free combination

Hong Kong Chest Service1989, 1991^{27 28}

2 months Ritafer FDC + streptomycin (n=314) vs free combination (n=313)

Nausea and vomiting 38 vs39% Problems taking pills 1% vs 5% (p<0.05) Regularly brought drink to help take pills 32% vs 45% (p<0.01)

Directly observed therapy High attendance in both groups

Small advantage of FDC in acceptability to patients

RCT

FOLLOWED BY ISONIAZID, RIFAMPICIN AND STREPTOMYCIN± PYRAZINAMIDE FOR 2 MONTHS and isoniazid, rifampicin ± pyrazinamide for months

Adverse reactions: similar in FDC and free
Culture negativity At 1 month 68 vs 58%, 65 vs 60%, 73 vs 66%, 59 vs 53% At 2 months96 vs 93%, 90 vs 86%, 92 vs 93%, 89 vs 88%groups
Relapse during 30 months from end of treatment: 5% vs 3.5% (p>0.05)

No differences shown in culture negativity at 1or 2 months, or in relapse rate

Zhang 1996, China²⁹

2 months Ritafer FDC + 4 months Rifinah FDC (n=104)
vs free combination (105)

Strong patient preference for FDC (p<0.01)
Strong physician preference for FDC (p<0.01)
Strong pharmacist/administrator preference for FDC (p<0.01)
Adverse reactions: 12% vs 16%

Directly observed therapy
High attendance in both groups

Equal efficacy and tolerability of FDCs
Advantage in acceptability by patients, physicians, pharmacists and administrators

Culture negativity At 2 months 99 vs 96%(p>0.05) At 6 months99 vs 98%(p>0.05)
Relapse during 18 months from end of treatment 2% vs 2%

Malaria

No trials found

HIV/AIDS

Eron 2000, US¹³

Combivir FDC bd with an FDA approved protease inhibitor (n=110)

Treatment failure (viral load) 3.6 vs 7.1%

Self-reported missed doses (diary cards) :>98% compliance for both groups.

Powered to show only non-inferiority of clinical outcomes for FDC so can’t demonstrate a benefit.

RCT

vs Lamivudine 150 bd, Zidovudine 200 tid with an FDA approved protease inhibitor (n=113)

Absolute difference = 3.5% (-2.4% to9.3%) (p=0.26)

Less missed doses of L/Z at: At 8 weeks (p=0.007) At 16 weeks (p= 0.046)

Clear improvement in adherence but too short in duration to show and effect on adherence to long-term treatment

for 16 weeks

Change in CD4+: Treatment difference = 5.9 (-15.8 to 27.6) cells/litre (p=0.59)

Adherence questionnaire: Better scheduling and timing scores At 8 weeks (p= <0.001) At 16 weeks (p=0.022) Better total scores At 8 weeks (p=0.002) At 16 weeks (p=0.020)

(Combivir= Lamivudine 150/Zidovudine 300)

Rozenbaum 1988, France³⁴

Combivir FDC bd (n= 35)

Log 10 HIV RNA (median change from baseline) 1.26 -vs-1.29 copies/ml

Abstract only available

RCT

vs free combination (n= 40)

CD4 (median change form baseline 34 vs 38 cells/mm³)

Equivalent antiviral activity and same safety profile in both groups

for 12 weeks

Tolerance: 1 vs 4 pts with adverse events

Unit-of-use packaging vs free combinations

Tuberculosis

No trials found

Malaria

Yeboah-Antwi 2001, Ghana¹⁵

Chloroquine and paracetamol pre-packaged in unit doses (n=3 health centres; 314 patients)

Compliant for 3 days: Tablets 82.0% vs 60.5% Abs diff = 21.5% (11.8-31) (p< 0.001)

Precision of estimates not adjusted for cluster design

Cluster RCT

vs same drugs in usual presentation (n=3 health centres; 340 patients)

Syrup 54.7% vs 32.6% Abs diff = 22.1% (8.3-36) (p< 0.001)

Precision of estimates not adjusted for cluster design

Total 72.1 vs 49.8% Abs diff = 22.3% (14.1-31) (p< 0.001)

50% reduction in cost of treatment, 50% reduction in waiting time at clinic

Qingjun 1998 China³⁰

Chloroquine 4 tablets on day 1, 3 tablets on days 2 and 3 in blister pack + Primiquine 3 tablets on days 1-8 in blister pack Oral and written instructions (Phase 1: n= 161, Phase 2 : n=138)

All patients smear-negative and symptom free following treatment

Phase 1: Compliance 97 vs 83% (p<0.01)

Marked improvement in compliance with blister packaging and written instructions (as well as oral instructions for both groups). Most of the improvement in compliance was related to the preservation of the medication in the blister packs. No difference in efficacy shown

RCT

Of 27 non-compliant in control group, 16 had lost the medication or it had dissolved or crumbled in the paper envelope

vs free combination in paper envelope, oral instructions only (Phase 1: n=163, Phase 2: n=134)

Phase 2: Compliance 97 vs 81% (p< 0.001)

HIV/AIDS

No trials found

FDCs vs unit-of-use packaging

No trials found