The effect of drug packaging on patients' compliance with treatment for Plasmodium vivax malaria in China.

The effect of drug packaging on patients' compliance with treatment for Plasmodium vivax malaria in China Li Qingjun,l 4 Duan Jihui,2A Tang Laiyi,34 Zhang Xiangjun,4 Liang Jun,4 A. Hay,5 S. Shires,5 & V. Navaratnam6 A study conducted in 1994 showed that the use of blister packs containing antimalarial drugs significantly increased patients' compliance, compared with traditional means of dispensing drugs in a paper envelope. The present study assessed patients' compliance and compared the difference between 3-day chloroquine and 8-day primaquine courses of treatment for vivax malaria. The level of real compliance was determined by marking the drugs with phenobarbital, and measuring its level in the blood following treatrnent. The results show that blister packaging significantly improved patients' compliance (P <0.001) over traditional means of dispensing antimalarial drugs; there was no difference in treatment compliance between 3-day and 8-day courses when the drugs were in blister packs. However, with ordinary packaging the treatment compliance rate for an 8-day course was significantly less than for a 3-day course (P < 0.05). Introduction Hunan Province in south-central China comprises 14 prefectures and 125 counties with a population of 64 million. Since the start of the malaria control pro- gramme in this province in the 1950s, there have been three Plasmodiuim vivax malaria outbreaks with annual incidences of 1560 per 100000 in 1955, 1690 per 100000 in 1962, and 1323 per 100000 in 1971. Despite a comprehensive malaria control pro- gramme, the incidence since 1986 has risen and, for counties in the south and south-west of the proNince, is five times higher than the provincial average. Chloroquine and primaquine are produced in China for the treatment of P. vivax malaria, and are generally dispensed by health workers from 1000- tablet bottles or tins. The current treatment regimen for adults is 4 chloroquine tablets (each containina 0.25mg chloroquine phosphate) on day 1, 3 tablets on days 2 and 3, and 3 tablets of 13.2mg primaquine daily from day 1 to 8. The curative dosage of each drug is dispensed by a health worker or pharmacist IDirector. 2 Professor of Parasitology. 3Professor of Epidemiology. 4Health and Epidemic Prevention Station Hunan Province, Changsha, Hunan, China. 5 Division of Clinical Sciences, School of Medicine, Old Medical School, University of Leeds, Leeds, England. e Director, Centre for Drug Research, Universiti Sains Malaysia, Penang, Malaysia. Requests for reprints should be sent to Dr M. Gomes at the address shown on p. 9. Reprnt No. 5847 from the container in a paper envelope or small pa- per bag, and given to the patient without any identi- fication of the drugs or written directions on their use. The doctor's or health worker's oral instructions to the patient on the number of tablets to be taken and for how many days are brief and do not include any advice or health education. Treatment is often incomplete due to its discontinuation once the symp- toms have disappeared, or to disintegration or loss of the tablets. Sometimes patients take an excessive dose so as not to risk loss of the drug, and can expe- rience unpleasant side-effects. Over 20 years ago, compliance was assured by delivering drugs on a daily basis to patients with direct observation of therapy. While clearly effec- tive. this was unsustainable owing to the need for large numbers of health workers. To improve com- pliance and use of antimalarial drugs, a pilot inter- vention study was undertaken to determine the effects of new packaging of drugs and written instructions on compliance. The specific objectives of the study were as follows: - To compare the compliance of patients in completing a full course of antimalarial treat- ment; a control group received medication in pa- per envelopes and a second group were given the improved packaging with simple, written instruc- tions on their use. - To determine the possible reasons for non- compliance in both groups and whether they were related to the type of packaging and instruc- tions on use. Builetin of the World Health Organization, 1998, 76 (Suppl. 1): 21-27 0 World Health Organlzation 1998 21 U QingJun et al. - To evaluate the impact of the new packaging and of providing simple, written instructions with the medication. - To determine the difference between reported compliance (Phase-I study, based on only pa- tients' reports) and true compliance (Phase-Il study, including use of a pharmacological marker inserted with each antimalarial tablet). Materials and methods Blister packs. The tablets of chloroquine and primaquine for the study were provided in two packs, one for chloroquine (with daily doses sepa- rated for three consecutive days) and the second for primaquine (with daily doses of 3 tablets for a total of 8 days). The number 8, according to Chinese be- liefs, brings good luck and it was considered that a course of primaquine over 8 days would enhance compliance more than one for 14 days. Detailed oral instructions were given to the patients on the correct dosage and the need to complete the treatment. The new blister pack was hermetically sealed, separately for each day's supply of tablets, with the name of the drug on the back of the pack. The blister packs were inserted into boxes, with the drug name on the front and dosage directions and precautions on the back; clear instructions were given on a leaflet inside the box. Each pack measured 8 X 5.5cm and contained 10 tablets of chloroquine for use over 3 days, or 24 tablets of primaquine for 8 days. Study sites. For the phase I study, which was con- ducted during the malaria season in 1994, the Health and Epidemic Prevention Stations in five counties of Hunan Province were selected as the study site. In phase II, the Health and Epidemic Prevention Stations in seven counties implemented the study in the malaria season from July till October 1996. The choice of epidemic station in the counties was made so that a sufficient number of patients could be enrolled in the study period, which would ensure that the study was carried out as designed. The quali- fications of the staff at these stations were high, and each station had a good working relationship with the provincial authorities. During the study period, millions of people from Hunan Province migrated to Guangdong and Hainan Provinces where a high level of economic development was taking place. Most workers were engaged in mining and road and house constructions in areas of high malaria endemicity. Malaria is com- mon among the migrant. nonimmune populations and when these persons return to their homes in Hunan, the disease is introduced, causing epidemics of imported malaria. Patients. The patients in the study were consenting adults (aged >15 years) with confirmed, slide- positive vivax malaria; they were ambulatory, had no major clinical symptoms requiring hospitaliza- tion, and had not received any malaria treatment during the previous 6 months. On selection at one of the clinic sites, a patient was randomly assigned either to the new drug packaging group or to the control group prior to receiving medication. Upon entry to the study, basic data were recorded on each patient, including age, sex, occupation, suspected onse't of infection, clinical symptoms, laboratory test results, and date and time of treatment. When the drug regimen was completed in phase I and phase II, information was solicited on when the first and last doses of the drug were taken, the total number of tablets taken, and on compliance and non-compliance. Phase J. As mentioned above, this phase was carried out in five counties of Hunan Province. Patients in the control group were prescribed medication with oral instructions on their use; the antimalarial tablets (chloroquine and primaquine) were put into a paper envelope and given to the patients. Patients in the intervention group were prescribed medication with identical oral instructions; the antimalarial tablets were given in two separate blister packs, marked with appropriate daily dosages and containing de- tailed written instructions. The advice given by the doctor to both the control and the intervention groups was the same regarding malaria and treat- ment. However, each patient in the intervention group was asked to open the blister pack and read the instructions for treatment found inside. When they had finished reading the instructions, they were asked if they understood what they had just read. Patients in the control group were asked if they had understood the doctor's oral instructions. If the pa- tient indicated a poor understanding, the instructions for treatment were repeated for both groups. The sample size was determined using an expec- tation of 15% noncompliance in the control group and an expected 10% improvement in compliance (or 15% noncompliance) in the intervention group. With a desired confidence level of 95%, a power of 80%, and a ratio of 1:1 in the control and interven- tion groups, the sample size was determined to be 318 persons - 159 in the control group and 159 in the intervention group. The choice of 15% noncom- pliance and an expected improvement in a compli- ance of 10% was conservative as was also the sample size. WHO Bulletin OMS. Vol 76 (Suppl. 1) 199822 Effect of drug packaging on compliance with antimalarial treatment On day 9 (after the start of treatment), each patient was either visited at home by a doctor in the study team or patients returned to the Epidemic Station for follow-up. All patients were retested for the presence of malaria parasites. In addition, a questionnaire was administered regarding the pa- tient's compliance, whether the medication had been taken according to the doctor's advice, the method of taking the medicine, the duration of treatment, and any causes of an incomplete course of treatment. Phase It. This was a randomized study, which was carried out in seven counties of Hunan Province. Upon entering any one of the seven clinic sites, a patient was randomly assigned to one of the follow- ing treatment groups: - Intervention group 1: Blister pack with phenobarbital-marked chloroquine plus blister pack with unmarked primaquine. - Intervention group 2: Blister pack with un- marked chloroquine plus blister pack with phenobarbital-marked primaquine. - Control group 3: Normal package with phenobarbital-marked chloroquine plus un- marked primaquine. - Control group 4: Normal package with un- marked chloroquine plus phenobarbital-marked primaquine. To obtain the marked drugs, chloroguine and primaquine tablets were each reground, marked with phenobarbital, and put into capsules by the Centre for Drug Research, Universiti Sains Malaysia. The capsules of chloroquine were white, as were the chloroquine tablets; the primaquine capsules were dark pink, while the tablets were red. All the pheno- barbital was obtained from the same source - the laboratory of the Universiti Sains Malaysia. The marked capsules and unmarked tablets for the intervention groups were blister packaged, and boxed with instructions on proper drug administra- tion. These same drugs were also available in bulk for administration to the non-blister-pack groups. The marked chloroquine capsule contained the nor- mal daily dose (150mg chloroquine phosphate + lmg phenobarbital). The marked primaquine cap- sule contained the daily dose (39mg primaquine phosphate + 2mg phenobarbital). The unmarked chloroquine was the standard single dose tablet and the normal daily dose of unmarked primaquine was contained in 3 tablets. Groups 1 and 3 were required to take a total of 10 capsules of marked chloroquine (4 capsules on dav 1 followed by 3 capsules on each of days 2 and 3) and 24 tablets total of primaquine over 8 days (3 tablets daily for 8 days). Groups 2 and 4 were re- quired to take 8 capsules of marked primaquine (1 capsule daily for 8 days) and 10 tablets of unmarked chloroquine (4 tablets on day 1 followed by 3 tablets on each of days 2 and 3). Groups 1 and 3 therefore received the same drug formulations, with group 1 receiving the blister pack; groups 2 and 4 received the same drug formulations, with group 2 receiving the blister pack. Each patient was either visited at home by a doctor in the study team or returned to the county Epidemic Station for follow-up. This was done on the fourth day for the marked chloroquine interven- tion and control groups (groups 1 and 3), and on the ninth day for the marked primaquine intervention and control groups (groups 2 and 4). Patients were asked whether they had completed the course of treatment, and 4ml blood was drawn from each pa- tient. The blood was separated by centrifuge, and the plasma was stored in two test tubes at -25 IC. One sample was subsequently sent to the Centre for Drug Research, Universiti Sains Malaysia, Penang, Malay- sia, and the other to the Department of Pathological Sciences/Chemical Pathology, School of Medicine, Leeds University, England, for phenobarbital de- tection. Plasma samples were analysed using the methodology given in the paper "Initial evaluation of low-dose phenobarbital as an indicator of compli- ance with antimalarial drug treatment" in this Sup- plement of the Bulletin (see p. 67-73). A subgroup of 57 patients was followed up 1- 100 days after the treatment to determine any recru- descence and to assess, through interview, their acceptance and compliance with drugs in the blister packs. They were asked whether they were satisfied aith the new packaging or the paper envelopes, whether they had read the printed instructions in the box. and whether they had taken the medicine ac- cording to the given instructions. A total of 320 patients were to be recruited. However. the trial was stopped earlier than planned because an analysis when 272 vivax malaria patients had been recruited into the study in the seven sites revealed that the numbers required to show significance between the two groups had been reached. A total of 138 patients were assigned to the intervention group (62 received blister packs with phenobarbital-marked chloroquine and 76 received blister packs with phenobarbital-marked primaquine); 134 patients were assigned to the control group (70 received paper envelopes with phenobarbital-marked chloroquine and 64 received paper envelopes with phenobarbital-marked primaquine). WHO Bulletin OMS. Vol 76 (Suppl. 1) 1998 23 Li Qinglun et al. Table 1: Compliance, by county, In the Intervention and control groups Interventon group Control group County No. of cases No. who complied No. of cases No. who complied Shaoyang 38 33 (87)a 39 31 (80) Dongkou 39 39 (100) 41 35 (85) Longhui 13 13 (100) 16 15 (94) Daoxian 39 39 (100) 38 31 (82) Lanshan 32 32 (100) 29 24 (82) Total 161 156 (97) 163 136 (83) * Figures in parentheses are percentages. Table 2: Reasons given for noncompliance by patients in the Intervention and control groups Reason No. in intervention group No. in control group Doctors advice unclear 1 0 Misunderstood the doctor 1 3 Forgot to take medication 1 2 Just stopped taking medication 2 6 Medication was lost 0 8 Medication dissolved/crumbled 0 8 Total 5 27 Statistics. Wilcoxon-Mann-Whitney tests on phe- nobarbital level-to-dose ratios were performed using the Astute Programme for Microsoft Excel COOLL Software, University of Leeds. Results Phase 1. The 300 male and 24 female patients for evaluation were aged 16-63 years (mean, 31 years); three-quarters of the group were aged <38 years. Using patients' reports as a measure of compliance, Table 1 shows the extent of compliance in five coun- ties in the control and experimental groups: 97% of patients in the intervention group and 83% in the control group completed the proper treatment regi- men. This result is statistically significant P < 0.01 using the Kruskal-Wallis one-way analysis of vari- ance for data not normally distributed. If the num- ber of lost or spoilt medication is removed from the analysis, the result is not statistically significant (P = 0.13). From data (not presented here) compli- ance was not related to patient's age, occupation or education, or gastrointestinal distress - defined as nausea, vomiting, abdominal pain or diarrhoea. The reasons for not completing the treatment are given in Table 2. Following the treatment regimen, patients in both the control and intervention groups were retested for malaria by blood smear. All were nega- tive and asymptomatic. From data (not presented here) for those who reported that they had failed to complete the course, the majority took at least their full dose of chloroquine whisi would cure their symptoms and parasitaemia. Although the course of primaquine in the non-compliant group was incom- plete, all except one took at least 5 days of prima- quine which made the chances of relapse remote. Phase H. A total of 272 vivax malaria patients were enrolled in the study, ranging in age from 11 years to 67 years, the majority being aged 20-40 years; most patients were male. Of these, 138 were assigned to the intervention group (62 received blister packs with marked chloroquine, and 76 received blister packs with marked primaquine); 134 patients were assigned to the control group (70 receiving marked chloroquine and 64 marked primaquine). The overall treatment compliance for patients who received blister packs (combined chloroquine and primaquine) was 97.1% (134/138). The compli- WHO Bulletin OMS. Vol 76 (Suppl. 1) 199824 Effect of drug packaging on compliance with antimalarial treatment Table 3: Comparison of compliance rates between 3-day (chloroquine) and 8-day (primaquine) treatments Intervention group (blister pack) Control group (paper envelope) Treatment No. of cases No. who complied No. of cases No. who complied Marked chloroquine 31 29 (93.5)a 35 30 (85.7) Unmarked chloroquine 38 38 (100) 32 29 (90.6) 3-day compliance 69 67 (97.1) 67 59 (88.1) Marked primaquine 38 36 (94.7) 32 23 (71.9) Unmarked primaquine 31 31 (100) 35 26 (74.3) 8-day compliance 69 67 (97-1) 67 49 (73-1) Treatment compliance 138 134 (97.1) 134 108 (80.5) - Figures in parentheses are percentages. Table 4: Phenobarbital LDRs (level-to-dose ratios) In intervention (blister packs) and control (old packaging) groups for chloroquine and primaquine Chloroquine Primaquine Blister packs Old packaging Blister packs Old packaging (n=36) (n= 23) (n= 31) (n= 34) Minimum 1.2 1.99 2.29 3.74 25th percentile 3.26 2.92 5.27 6.52 Median 3.86 3.46 7.25 7.12 75th percentile 4.86 4.8 9.88 9.19 Maximum 18.85 30.9 19.99 30.24 w 375 453 569.5 484.5 U 1119 651 980.5 1164.5 Corrected for ties: z -0.6061 -0.6062 -0.5582 -0.604 P 0.5444 0.5444 0.5767 0.5444 ance for patients whose medication was in paper envelopes (combined chloroquine and primaquine) was 80.5% (108/134). The difference is statistically significant (P < 0.001). A comparison of 3-day and 8-day treatments showed that the blister-pack group had 97.1% (67/ 69) compliance with 3 days of chloroquine and 97.1% (67/69) compliance with 8 days ofprimaquine. For the control group, 88.1% (59/67) completed the 3-day treatment, and 73.1% (49/67) completed the 8- day treatment (Table 3). The difference between the two groups for both the 3-day and 8-day regimens is statistically significant (P < 0.05). The rate of noncompliance and reasons given for noncompliance are similar in both studies. The compliance (blister pack vs paper package) in 1994 was 96.9% and 83.4% and in 1996, 97.1% and 80.5%. respectively. The reasons included drugs got lost or dissolved, or the patient simply forgot. Among 57 patients who received a second follow-up visit 1-100 days after the treatment, 28 were from the blister pack group and 29 from the control group. One patient from the control group had a relapse; he had not completed his treatment and recrudesced 88 days after the initial diagnosis. Of 55 patients who responded, 48 (87.3%) preferred the blister pack because they were more hygienic, convenient to carry, and more difficult to lose. Of 26 patients who had used the new packaging, 22 (88.7%) had read the instructions, 21 (80.8%) under- stood the instructions, and 14 (53.9%) took the medication exactly according to the instructions. The overall phenobarbital level-to-dose ratios (LDRs) for the intervention and control groups are given in Table 4. The differences between the two groups are not significant. Discussion Tlhe cure rate of P. vivax malaria with 8 days of chloroquine and primaquine treatmnent is 99.2% (1). This therapy is wvidely practised in China, and consid- ered safe and effective (2). However, because of the WHO Bulletin OMS. Vol 76 (Suppl. 1) 1998 25 U Qinglun et al. bulk packaging of chloroquine and primaquine, their distribution is inconvenient for both doctors and patients. The result is inefficiency due to misunder- standing the necessity for completing the treatment, and possibly increased recrudescence rates (3). Compliance with antimalarial treatment ap- pears to be affected by the type of packaging and instructions given on dosage and treatment. This observation does not necessarily support the conclu- sion that the blister pack is the best method of distri- bution, because poor compliance could be due to poor packaging. Paper has limited properties for protecting water-soluble medication, and 16 out of 27 patients in the control group in phase I were non- compliant as a result of losing the medication or tablet disintegration. None of the patients in the in- tervention groups were noncompliant for these rea- sons. Indeed, if lost or disintegrating tablets are discounted because they are beyond the control of the patient, there is no statistical difference in com- pliance between the two groups although there was a difference between the aggregate number of patients in the intervention and control groups who were compliant. From the patients' perspective, the blister pack was treated with more care, and this may be reflected in a higher compliance rate. The rate of completing the treatment among blister-pack recipients was significantly higher than the paper package control group. This holds true for both a 3-day and 8-day regimen. As some of the reasons given for noncompliance were related to the poor paper packaging leading to crumbling and loss of medication, the blister pack resolves these deficiencies. The above results using phenobarbital as a marker of compliance are inconclusive. The LDRs in all groups were considerably higher than ex- pected, possibly 'because these patients had con- sumed other medications containing.low doses of phenobarbital. Comparison of the chloroquine and primaquine LDRs, excluding.patients with the high values, was also not significant, as-was a comparison of the control population's LDRs with the levels ob- tained in the Thai reference study (4, see pp. 59-66 in this Supplement). To assess the value of using phenobarbital as a marker of compliance, data from this study should be compared with data from a ref- erence group of good compliers, who would have consumed the phenobarbital under observation. Acknowledgements We are grateful to the following health units and staff for their collaboration: Health and Epidemic Prevention Sta- tion of Hunan Province, Changsha: Li Zhengxing; Health and Epidemic Prevention Station of Dongkou County, Dongkou: Deng Qune; Health and Epidemic Prevention Station of Shaoxang County, Shaoxang: Uu Jirong and Wang Zhaohui; Health and Epidemic Prevention Station of Daoxian County, Daoxian: Uu Almin; Healfth and Epidemic Prevention Station of Guiyang County, Guiyang: Peng Hairong and Qiu Deming; Health and Epidemic Prevention Station of Yongzhou City, Yongzhou: Deng Tianyu; Health and Epidemic Prevention Station of Jianghua County, Jianghua: Wang Jianghao; Health and Epidemic Preven- tion Station of Leiyang, Leiyang: Uu Houzi; Centre for Drug Research, Universit Sains Malaysia: Dr S.M. Mansor. This study was supported by the UNDP/World Bank! WHO Special Programme for Research and Training in Tropical Diseases (ID:940883). Resume Effet du conditionnement des m6dlcaments sur I'observance du traitement du paludisme & Plasmodium vivax en Chine Une 6tude r6alis6e en 1994 a montre que le conditionnement des antipaludiques sous plaquette thermoform6e ("blister") am6liore sensiblement l'observance du traitement par rapport A la dis- pensation traditionnelle dans une enveloppe en papier. La pr6sente etude evalue l'observance du traitement et compare la difference liee au conditionnement entre le traitement de 3 jours et le traitement de 8 jours par la primaquine dans le paludisme A vivax. Le taux effectif d'observance a ete determine par marquage des m6dicaments avec du phenobarbital, dont a mesure le taux sanguin & I'issue du traitement. Les resultats montrent que le.conditionnement sous plaquette thermoforme'e augmente de fa9on significative l'observance (p < 0,001) par rapport aux moyens traditionnels de dispensation; aucune diff6rence d'observance n'a et6 trouv6e entre le traitement de 3 jours et celui de 8 jours lorsque les medicaments 6taient present6s sous plaquette thermoformee. Avec la presentation habituelle, le taux d'observance etait significativement plus faible pour le traitement de 8 jours que pour celui de 3 jours (p < 0,05). References 1. Huei Bailin et al. The radical efficacy of pyronaridine- primaquine combination in tertian malaria. Joumal of parasitology and parasitic diseases, 1984, 2(4): 262 (in Chinese, summary in English). 2. Zhou. Malaria control and research in China. 1st edi- tion. Beijing, People's Health Press, 1991: 157-207. 26 WHO Bulletin OMS. Vol 76 (Suppl. 1) 1998 Effect of drug packaging on compliance with antimalarial treatment 3. Wang Ke'an et al. Relapses and indirect fluorescent antibody level in vivax malaria patients after chloroquine and chloroquine-primaquine treatment. Joumal of parasitology and parasitic diseases, 1984, 2(4): 226-227 (in Chinese, summary in English). 4. Fungladda W et al. Compliance with artesunate and quinine + tetracycline treatment of uncomplicated falciparum malara in Thailand. Bulletin of the World Health Organization, 1998, 76 (Supplement 1): 59- 66. WHO Bulletn OMS. Vol 76 (Suppl. 1) 1998 27