8

© 2006 International Society of Travel Medicine, 1195-1982Journal of Travel Medicine, Volume 13, Issue 1, 2006, 8–14

Do Travelers Really Take Their Mefl oquine Malaria Chemoprophylaxis? Estimation of Adherence by an Electronic Pillbox

Pierre Landry , MD , * Danila Iorillo, RPh , † Roger Darioli , MD , † Michel Burnier , MD , † and Blaise Genton , MD , PhD * * Travel Clinic , Medical Outpatient Clinic, University of Lausanne, Lausanne, Switzerland ; † Medical Outpatient Clinic, University of Lausanne, Lausanne, Switzerland

DOI: 10.1111/j.1708-8305.2006.00005.x

Background. Nonadherence to chemoprophylaxis could explain why some travelers get malaria. Adherence is notori-ously diffi cult to assess, and most studies have been conducted using questionnaires. This study aims at assessing con-tinuous adherence more accurately with the help of an electronic pillbox. Methods. Adult travelers to sub-Saharan Africa had to fi ll a questionnaire on demographic and travel data, drug intake, and adverse events. They received oral and written information about malaria and mefl oquine prophylaxis and a Medica-tion Event Monitoring System (MEMS ® , Aardex ® , Zug, Switzerland), ie, a bottle closed with a cap containing a micro-processor recording date and time of all openings, fi lled with the exact number of mefl oquine 250 mg tablets (Lariam ® , Roche Reinach, Switzerland) . The MEMS ® was returned with the questionnaire after completion of chemoprophylaxis. Results. According to the MEMS ® , only 26 of 81 travelers (32.1%) took all the doses at the expected date, another 8 (9.9%) did so but starting late with the fi rst dose, and 19 others (23.5%) took all the pills but with intervals of �1 day from the right date. Another eight (9.9%) took all the pills but in a random way. The remaining 20 travelers (24.7%) missed some doses, mainly after return. Strict adherence as assessed by electronic monitoring was therefore lower than adher-ence measured by questionnaire (32.1% vs 48% in taking all the tablets on the right day). There was no difference between the two methods when a broader defi nition of adherence was applied [taking all the tablets on the right day (�1 day); 53/81 (65.4%)], but the MEMS ® showed that some answers to the questionnaire were not reliable. Conclusion. The use of electronic pillboxes confi rms the low adherence of travelers to mefl oquine chemoprophylaxis in spite of extensive information about the disease and its prevention. Electronic assessment of pill taking, for the fi rst time applied to malaria chemoprophylaxis, gives new insights into the real adherence of travelers.

Regular chemoprophylaxis combined with protec-tive measures against mosquito bites can largely prevent malaria infection and disease. 1 – 4 Imported malaria cases are increasing. Likely explanations are the higher numbers of people traveling to

endemic areas and development of drug-resistant malaria. Nonadherence to recommended malaria- preventive schedules might be one potential ex-planation for the apparent reduced effi cacy of antimalarial drugs. 4,5

Studies in returning travelers show that most cases of malaria occur in those taking no protective measures and not using any prophylaxis or with the wrong drug or dosage. 1,6 – 8 Adherence is a complex concept mixing understanding of risks, understand-ing of protective/curative interventions, acceptance of these measures as relevant for oneself, and their application in spite of discomfort. 9

Previous studies assessing travelers ’ adherence to chemoprophylaxis used questionnaires 3,5,10 – 20

Preliminary results were presented as a poster [Iorillo D, Landry P, Darioli R, et al. Electronic monitoring of travelers ’ adherence to mefl oquine malaria chemopro-phylaxis (Abstract C28). Sixth Conference of the Inter-national Society of Travel Medicine; 1999 Jun 6 – 10; Montreal, Canada].

Corresponding Author: Pierre Landry, MD, Place Pury 9, CH-2000 Neuchâtel, Switzerland.

Estimation of Adherence by an Electronic Pillbox 9

J Travel Med 2006; 13: 8–14

with the usual limitations for this method of investi-gation, such as forgetting the precise date and time of drug intake, especially in retrospective question-naires, or overreporting of the correct regimen in an effort to please the investigators or to hide one ’ s failure. This has been put into light in the discrep-ancies between blood levels of drug and question-naire reporting. 21

A new electronic device able to record date and time of all bottle openings is available to monitor adherence with any treatment more accurately than by questionnaire. 22,23 This method has been mainly applied to drugs taken on a daily basis (but no anti-malarial drug to our knowledge). 24 – 26 Data for a weekly dosing regimen such as mefl oquine chemo-prophylaxis are even fewer. 27

Objectives

1. To evaluate travelers ’ adherence to malaria che-moprophylaxis with weekly mefl oquine 250 mg (Lariam ® , Roche, Basel, Switzerland).

2. To compare adherence as assessed by electronic monitoring with adherence as assessed by stan-dard questionnaire.

Subjects and Methods

Subjects Between March and April 1997, every person older than 18 years, traveling to Africa for less than 3 months and consulting our Travel Clinic, Medi-cal Outpatient Clinic, University of Lausanne, Switzerland, was asked to take part in our prospec-tive study. Pregnant women and people with con-traindication to mefl oquine were excluded. Only one person per couple was included. The protocol was approved by the institutional review board.

Methods Subjects received extensive oral and written infor-mation on malaria, its way of transmission, protec-tive measures against mosquitoes, mefl oquine chemoprophylaxis, and symptoms of the disease. They were asked to fi ll an open questionnaire for demographics, medical history, current medical treatment, and travel characteristics before their trip and to complete information on the use of pro-tective measures against exposure, date and time of drug intake, and reasons for missing doses, as well as date, type, and severity of adverse events or any ill-ness related to the journey after they had fi nished the chemoprophylaxis.

Each one received a Medication Event Monitor-ing System (MEMS ® , Aardex ® , Zug, Switzerland), ie, a bottle closed with a cap containing a micropro-cessor recording date and time of all openings (see Figure 1 ) and fi lled with the exact number of mefl oquine 250 mg tablets required for the journey. To avoid problems with the custom authorities and because of their sensitivity to humidity, tablets were kept in their original blisters. The travelers were made aware of the fact that the MEMS ® would re-cord all bottle openings and were instructed not to open it except to take the prescribed regimen. The MEMS ® and the questionnaire were to be returned as soon as the last dose had been taken. A remainder letter or phone call was done to trace nonreturned MEMS ® .

Malaria Chemoprophylaxis Regimen The malaria chemoprophylaxis was prescribed according to the Swiss recommendations of the Swiss Offi ce of Public Health, ie, one tablet of mefl oquine 250 mg once a week, starting 1 week before departure and up to 4 weeks after return. Subjects visiting the Travel Clinic less than 1 week before departure were asked to take one tablet on the day of their visit and the second one, 5 days after, before resuming the weekly recom-mended schedule. Dosing for people more than 90 kg body weight was increased to one and half tablet (375 mg) per week.

Defi nition of Adherence In this study, complete adherence to prophylaxis was defi ned as the appropriate taking of all tablets of mefl oquine (including 4 weeks after return) at the right date and acceptable adherence as taking of all the tablets at the right date (�1 day), even if starting less than 7 days before the journey.

Data Analysis MEMS ® data were analyzed with CSS software (Aardex ® ) and Microsoft ® Excel 97 SR-1 (Micro-soft Corp., Redmond, WA, USA). Data from the questionnaires and the MEMS ® were analyzed us-ing Epi Info, Version 6.04b software (Centers for Disease Control and Prevention, Atlanta, GA, USA, and World Health Organization, Geneva, Switzerland).

Results

From a total of 104 subjects fulfi lling the inclusion criteria, 4 refused to enter the study. Of the 100 enrolled, 4 canceled their trip, 1 changed the type

10 Landry et al.

J Travel Med 2006; 13: 8–14

of chemoprophylaxis, 1 had a journey longer than 3 months, 11 did not return their MEMS ® and/or questionnaire, and 2 persons had MEMS ® with technical failure, leaving 81 subjects for analysis. Of these, 18 (22.2%) had to be recalled to send their material back.

Assessment of Adherence by MEMS ® According to the MEMS ® , 61 of 81 (75.3%) trav-elers took all the tablets, but only 26 of 81 (32.1%)

travelers took all the recommended doses at the expected date, as prescribed ( Figure 2 ). Another 8 of 81 (9.9%) travelers did so but started late with the fi rst dose (less than 7 days before the depar-ture), and 19 of 81 (23.5%) took all the tablets but 1 day early or 1 day late on at least one occasion, and 8 (9.9%) took them all but in a random way. Therefore, 20 of 81 (24.7%) missed one or more doses.

Demographics and Travel Characteristics According to the Level of Adherence Table 1 gives the demographic and travel charac-teristics of the travelers according to their level of adherence. There was no signifi cant difference in terms of gender, length of travel, time before de-parture, destination, or purpose of travel. Travel-ers with an acceptable adherence were slightly older than nonadherent ones (37 years vs 31.5 years, p = 0.06), and completely adherent ones even more so (38 years). This latter group was traveling for a shorter period [mean 22.7 days (SD 18.2) vs mean 24.9 days (SD 13.5), p = 0.04] when compared to all other travelers. Having taken chemoprophylaxis on a previous trip had no infl u-ence on adherence, nor had the destination. Trav-elers for business or humanitarian purposes were more often adherent, but numbers were small.

Figure 1 Medication Event Monitoring System, Aardex ® .

Figure 2 Assessment of adherence by Medication Event Monitoring System (MEMS ® ). Note: Nonadherence in gray.

Estimation of Adherence by an Electronic Pillbox 11

J Travel Med 2006; 13: 8–14

Recalls by phone or written notice to recover questionnaires and MEMS ® were signifi cantly more common among nonadherent travelers [13 of 28 (46%) nonadherent were recalled vs 5 of 53 (9.4%) adherent ones, (RR 0.33; 95% CI 0.2 – 0.56)]. The theoretical median length of mefl o-quine intake for the whole group was 8 weeks (range, 7 – 17 weeks) .

Twenty-six travelers (32%) reported some mi-nor health problem prior to their travel, and 14 (17%) where taking some medication (seven times antihistaminic or antiasthma). These characteris-tics had no infl uence on adherence.

Adverse Events The following adverse events were reported by 34 of 81 (42%) travelers: gastrointestinal (nausea, vomiting; n = 16), anxiety or depression ( n = 7), dizziness ( n = 7), sleeping disorder ( n = 4), tiredness ( n = 4), headache ( n = 4), and other ( n = 4). Half of them (16/34) experienced adverse events after one or two tablets already. The adverse events were con-sidered as mild by 9 (27.3% of travelers reporting a dverse events), moderate by 11 (33.3%), and se-vere by 13 (39.4%). Travelers who took all their pills were slightly less likely to have suffered from adverse events [RR 0.48 (0.22 – 1.05), p = 0.06], but there was no statistical difference in the incidence of adverse events between adherent (complete or ac-ceptable) and nonadherent travelers.

Comparison between MEMS ® and Questionnaire Strict adherence as described above was lower when measured by electronic monitoring (32%) than by questionnaire (48%; Table 2 ). When a broader def-inition was used [taking all the pills at the right day (�1 day)], adherence by electronic monitoring was 65%. When asked if they had taken all the tablets, 62 travelers said that they had, but this was con-fi rmed in only 59 by the MEMS recordings. When asked if they had taken one or more tablets at an-other date than recommended, 26 travelers admit-ted it. This was correlated with the MEMS openings in 15 cases. Among the 11 others, 6 stopped com-pletely to take the tablets according to MEMS, leav-ing 5 cases whose MEMS were opened at the right date but stating in the questionnaire they had not. From the 53 travelers stating that they had not taken any pill on another date than recommended, 10 were found to have done so by MEMS. Therefore, in 21 cases (25.9%), the questionnaire and the MEMS gave discordant results. The questionnaire was not designed to ask for detailed timing of each intake; therefore, it could not provide some of the data given by MEMS in the Table 2 .

Discussion

This is the fi rst study assessing adherence to malaria chemoprophylaxis in travelers, using an electronic recording device.

Table 1 Demographic and travel characteristics according to the level of adherence

Adherent travelers

Complete adherence, Acceptable n = 26 adherence, n = 27

Nonadherent Total travelers, Difference adherent/ travelers, n = 28 n = 81 nonadherent

Gender M/F, n (%) 14/12 (53.9/46.1) 15/12 (55.6/44.4) 13/15 (46.4/53.6) 42/39 (51.9/48.1) p = 0.32 Age * (years) 38 (13.4) 35.9 (11.1) 31.6 (10.0) 35.1 (11.7) p = 0.06 Length of stay * (years) 22.7 (18.2) 24.2 (12.1) 25.5 (14.8) 24.2 (15.1) p = 0.23 Time before travel * (days)

26.4 (14.3) 20.9 (12.1) 22.1 (15.3) 23.1 (14.0) p = 0.49

Adverse events, n (%) 10 (38.5) 10 (37.0) 14 (50.0) 34 (42.0) p = 0.20 Destination East Africa 14 12 11 37 1.15 (0.84 – 1.57) † West Africa 6 11 10 27 0.94 (0.67 – 1.32) † Central Africa 5 3 4 12 1.02 (0.66 – 1.58) † South Africa 1 1 3 5 0.60 (0.20 – 1.76) † Main purpose Business 4 2 0 6 1.60 (1.34 – 1.90) † Humanitarian/ social 1 6 0 7 1.61 (1.35 – 1.92) † Tourism 15 10 19 44 0.75 (0.55 – 1.03) † Visit friends/ family 6 9 9 24 0.94 (0.65 – 1.34) †

* Continuous variables expressed as means (standard deviation). † Relative risk (95% confi dence intervals).

12 Landry et al.

J Travel Med 2006; 13: 8–14

According to MEMS ® , only 26 of 81 travelers (32%) took all their doses strictly following the rec-ommended regimen (ie, starting at the right date and taking it regularly). However, another eight (10%) did so with a delay in starting it, and 19 (23%) complied with chemoprophylaxis according to a �1 day schedule, which we judged to be an acceptable level of adherence. Indeed, for the latter group, and due to the long half-life of mefl oquine (25 days) and its kinetics, 28 – 30 we assume that taking the tablets 1 day before or after the scheduled day would hardly modify the blood level and hence protection.

Starting mefl oquine chemoprophylaxis only 1 week before a potential exposure as recommended in the guidelines could leave travelers with nonpro-tective blood levels in the fi rst weeks of their trip. This could have been a particular concern for the travelers who started to take their pill later than rec-ommended, even if they took them all and regularly. The real protection conferred by taking all the tab-lets but in a random way [8/81 travelers (10%)] is impossible to assess without monitoring mefl o-quine blood levels.

Previous studies aiming at assessing travelers ’ adherence to chemoprophylaxis with weekly me-fl oquine using oral or written retrospective ques-tionnaires gave adherence rates between 21 and 98%. 7,10 – 14 The interpretation of such a wide range

of results is diffi cult. Questionnaire assessment has been shown to overestimate adherence when com-pared to pill count or to drug blood levels. One im-portant reason for this is that the travelers do not remember when they take their tablets exactly. Also, they fear to displease the investigator or to appear as unreliable and therefore give what they think is an acceptable answer. Using an electronic device, re-cording openings should prove more accurate. In our study, the level of adherence was similar by both methods, but the comparison of both methods showed that in about 25%, the answers to questions of timing by questionnaire were unreliable. Eighteen travelers had to be recalled at least once to send the MEMS® and the questionnaire back after their trip, and they were more likely to be nonadherent. We presume that at least some of them fi lled their ques-tionnaire after return only and not during their trip, which could explain part of the discordant results be-tween both methods. We doubt that asking for more precise details about timing of each intake in the questionnaire could have given reliable information.

Both the methods are thought to have a remind-ing effect, the impact of which is not known. Our study was not designed to measure this impact, which could well have been of some importance as travelers were assessed by both methods at the same time. Therefore, our results like other studies prob-ably overestimate adherence, but the MEMS method gives some insights into the real behavior of travelers.

Nonadherence in taking any medication for prevention of hypothetical risks is higher than in taking medication against symptomatic diseases. This is probably even more so when on holiday. Ad-verse events are a known factor for poor adherence. In this study, about one-quarter of travelers did not take all the tablets, mostly after return, and the main reasons given were adverse events and forget -fulness, but there was no statistical signifi cance between adherent and nonadherent travelers. Non-adherence was linked to younger age and travel of longer duration. This has also been demonstrated in other studies. 12,16,19,20 Also, people traveling for business or short-term humanitarian reasons were more likely to be adherent than tourists, but the relative small number of participants did not allow for a complete study of the factors and characteris-tics leading to adherence or nonadherence.

A limitation of our study is the absence of another group assessed by questionnaire only, which could have given some more information regarding com-parative adherence between the two methods and about the reminding effect. Also, the cut blisters of

Table 2 Comparison between MEMS® and questionnaire

MEMS® Questionnaire

Full adherence (all the tablets at the right day)

26/81 (32.1%) 39/81 (48.2%)

Full adherence but started late

8/81 (9.9%) No data

All the tablets at the right day (�1)

45/81 (55.6%) No data

Took all the tablets but at various intervals

8/81 (9.9%) No data

All tablets (any time)

61/81 (75.3%) 62/81 (76.5%)

Did not take all the tablets

20/81 (24.7%) 19/81 (23.5%)

Missed one or more doses but resumed taking

3/81 (3.7%) No data

Stopped completely during the trip

3/81 (3.7%) No data

Stopped completely after the trip

14/81 (17.3%) 19/81 (23.5%)

Adverse events No data 34/81 (42.0%)

MEMS®, Medication Event Monitoring System.

Estimation of Adherence by an Electronic Pillbox 13

J Travel Med 2006; 13: 8–14

mefl oquine in the MEMS ® bottles could have lacked the reminder effect of a standard packet of tablets as to whether medication was taken or not.

Studies assessing compliance of weekly drug in-take with a MEMS ® are few. A trial comparing weekly versus daily fl uoxetine gave 87.8% adher-ence in the weekly group versus 79.0% ( p = 0.006) in the daily group. 27 Using questionnaires, adher-ence with weekly mefl oquine was 94.5% compared to 79.8% with daily chloroquin/proguanil. 12

A study of adherence to daily atovaquone pro-guanil HCL (Malarone ® , GSK, Muenchenbuchsee, Switzerland; was not yet marketed at the time of our study) or doxycycline (rarely given for destinations in Africa in our Travel Clinic) by MEMS ® could prove interesting.

Conclusion

The use of electronic pillboxes shows that adher-ence to mefl oquine chemoprophylaxis is too low in spite of a good oral and written information about the disease and its prevention. Most of the travelers who stopped the chemoprophylaxis did so after re-turn, and they tended to be younger and traveling for longer periods than those who were more ad-herent to recommendations. The electronic pill-boxes gave new insights in the way travelers were taking their pills, showing that very few did follow the recommendations strictly. It also showed that questionnaire answers about timing of intake were unreliable.

A cheap, well-tolerated, simple, and short-course chemoprophylaxis regimen is desirable to increase adherence in travelers, but unless it is a short-course regimen preferably before departure only, it is likely that adherence will remain unsatisfactory. In the mean time, travelers should be strongly advised to consult medical facilities in case of fever, especially if they stopped their chemoprophylaxis.

Acknowledgment

MEMS ® boxes and Lariam ® were provided by Roche Company.

Declaration of Interests

The authors state that they have no confl icts of interest.

References

1. Phillips-Howard PA , Radalewicz A , Mitchell J , Bradley DJ . Risk of malaria in British residents

returning from malarious areas . BMJ 1990 ; 300 : 499 – 503 .

2. Muehlberger N , Jelinek T , Schlipkoeter U , et al . Effectiveness of chemoprophylaxis and other deter-minants of malaria in travellers to Kenya . Trop Med Int Health 1998 ; 3 : 357 – 363 .

3. Adera T , Wolfe MS , McGuire-Rugh K , et al . Risk factors for malaria among expatriates living in Kampala, Uganda: the need for adherence to chemo-prophylactic regimens . Am J Trop Med Hyg 1995 ; 52 : 207 – 212 .

4. D ’ Acremont V , Landry P , Müller I , et al . Clinical and laboratory predictors of imported malaria: an aid to the medical decision making for returning travelers with fever . Am J Trop Med Hyg 2002 ; 66 : 481 – 486 .

5. Cobelens FGJ , Leentvaar-Kuijpers A . Compliance with malaria chemoprophylaxis and preventative measures against mosquito bites among Dutch trav-ellers . Trop Med Int Health 1997 ; 2 : 705 – 713 .

6. Reid AJC , Whitty CJM , Ayles HM , et al . Malaria at Christmas: risks of prophylaxis versus risks of malaria . BMJ 1998 ; 317 : 1506 – 1508 .

7. Zastrow KD , Dieckmann S , Schöneberg I . Malari-aeinschleppungen nach Deutschland 1989/1990 . Zbl Hyg 1993 ; 195 : 88 – 96 .

8. Centers for Disease Control and Prevention (CDC). Malaria deaths following inappropriate malaria che-moprophylaxis—United States, 2001 . MMWR Morb Mortal Wkly Rep 2001 ; 50 : 597 – 599 .

9. Kruse W . Patient compliance with drug treatment — new perspectives on an old problem . Clin Invest 1992 ; 70 : 163 – 166 .

10. Minodier P , Nassur A , Hassid S , et al . Antimosquito precautions and medical chemoprophylaxis in French children with malaria . J Travel Med 2003 ; 10 : 318 – 323 .

11. Carme B , Peguet C , Nevez G . Compliance with and tolerance of mefl oquine and chloroquine + proguanil malaria chemoprophylaxis in French short-term travellers to sub-Saharan Africa . Trop Med Int Health 1997 ; 10 : 953 – 956 .

12. Huzly D , Schönfeld C , Beuerle W , Bienzle U . Ma-laria chemoprophylaxis in German tourists: a pro-spective study on compliance and adverse reactions . J Travel Med 1996 ; 3 : 148 – 155 .

13. Held TK , Weinke T , Mansmann U , et al . Malaria prophylaxis: identifying risk groups for non-compli-ance . Q J Med 1994 ; 87 : 17 – 22 .

14. Phillips MA , Kass RB . User acceptability patterns for mefl oquine and doxycycline malaria chemopro-phylaxis . J Travel Med 1996 ; 3 : 40 – 45 .

15. Kollaritsch H , Wiedermann G . Compliance of Austrian tourists with prophylactic measures . Eur J Epidemiol 1992 ; 8 : 243 – 251 .

16. Schlagenhauf P , Steffen R , Lobel H , et al . Mefl o-quine tolerability during chemoprophylaxis: focus on adverse event assessments, stereochemistry and compliance . Trop Med Int Health 1996 ; 4 : 485 – 494 .

14 Landry et al.

J Travel Med 2006; 13: 8–14

17. Lobel HO , Phillips-Howard PA , Brandling-Bennett AD , et al . Malaria incidence and prevention among European and North American travellers to Kenya . Bull World Health Organ 1990 ; 68 : 209 – 215 .

18. Lobel HO , Campbell CC , Papaioanou M , Huong AY . Use of prophylaxis for malaria by American trav-elers to Africa and Haiti . JAMA 1987 ; 257 : 2626 – 2627 .

19. Lobel H , Baker MA , Graz FA , et al . Use of malaria prevention measures by North American and Euro-pean travelers to East Africa . J Travel Med 2001 ; 8 : 167 – 172 .

20. Laver SM , Wetzels J , Behrens RH . Knowledge of malaria, risk perception and compliance with che-moprophylaxis and personal and environmental pre-ventive measures in travelers exiting Zimbabwe from Harare and Victoria Fall International Airport . J Travel Med 2001 ; 8 : 298 – 303 .

21. Behrens RH , Taylor RB , Pryce DI , Low AS . Che-moprophylaxis compliance in travelers with malaria . J Travel Med 1998 ; 5 : 92 – 94 .

22. Urquart J . The electronic medication event monitor . Clin Pharmacokinet 1997 ; 32 : 345 – 356 .

23. Cramer JA , Mattson RH , Prevey ML , et al . How of-ten is medication taken as prescribed? A novel assess-ment technique . JAMA 1989 ; 261 : 3273 – 3277 .

24. Burnier M , Santschi V , Favrat B , Brunner HR . Mon-itoring compliance in resistant hypertension: an

important step in patient management . J Hypertens 2003 ; 21 ( Suppl 2 ): S37 – S42 .

25. Burnier M , Schneider MP , Chiolero A , et al . Elec-tronic compliance monitoring in resistant hyperten-sion: the basis for rational therapeutic decisions . J Hypertens 2001 ; 19 : 335 – 341 .

26. Schwed A , Fallab CL , Burnier M , et al . Electronic monitoring of compliance to lipid-lowering ther -apy in clinical practice . J Clin Pharmacol 1999 ; 39 : 402 – 409 .

27. Claxton A , de Klerk E , Parry M , et al . Patient com-pliance to new enteric-coated weekly formulation of fl uoxetine during continuation treatment of major depressive disorder . J Clin Psychiatry 2000 ; 61 : 928 – 932 .

28. Mimica I , Fry W , Eckert G , et al . Multiple dose ki-netic study of mefl oquine in healthy male volunteers . Chemotherapy 1983 ; 29 : 184 – 187 .

29. Hellgren U , Angel VH , Bergqvist Y , et al . Plasma concentrations of sulfadoxine-pyrimethamine and of mefl oquine during regular long term malaria prophylaxis . Trans R Soc Trop Med Hyg 1990 ; 84 : 46 – 49 .

30. Rombo L , Angel VH , Friman G , et al . Comparative tolerability and kinetics during long-term intake of lariam an fansidar for malaria prophylaxis in non-immune volunteers . Trop Med Parasitol 1993 ; 44 : 254 – 256 .