Estimated Occurrence of Tobacco, Alcohol, and Other Drug Use Among 12

Estimated occurrence of tobacco, alcohol, and other drug use among 12- to 18-year-old students in Panama: Results of Panama’s 1996 National Youth Survey on Alcohol and Drug Use

Gonzalo B. González,1 Miguel A. Cedeño,1 Marcel Penna,1 Luis Caris,2 Jorge Delva,3 and James C. Anthony 3

ABSTRACT This report provides the first epidemiological evidence on tobacco, alcohol, and other drug use among school students in Panama, using data from a student survey completed in 1996. Specifically, we examine sex, age, grade level, type of school, and urban-rural variations in the occurrence of tobacco, alcohol, and other drug use. Estimates of lifetime prevalence and past- year use of these products were obtained using data from Panama’s 1996 National Youth Sur- vey on Alcohol and Drug Use (n = 6 477). To account for the multistage sampling design of the survey, all estimates and respective standard errors are derived by the Taylor series approximation method using Epi Info 6.0 CSAMPLE software. In general, more males, more older students, and more students in higher grades have used licit and illicit drugs, even though male-female differences tend to be small. Public-private school differences and urban-rural trends vary depending on the drug. The findings of this study are discussed in relation to the epidemiology and prevention of drug use in Panama. Based on these data, we seek to provide information to be used by the Government of Panama in its planning for prevention programs directed toward students in Panamanian schools.

The Coalition of Panama [La Coali- the dimensions of youthful drug in- sent school-attending youths. The need ción de Panamá], organized by the First volvement in Panama, and to supply for such a survey was conveyed by the Lady of Panama, commissioned its information for the Coalition to use above coalition as a result of Panama’s Ministry of Public Health to conduct a in planning prevention programs di- drug surveillance system, which had nationwide survey of school students rected toward students in the coun- shown an increase in drug use and in order to provide new evidence on try’s schools. The purpose of this re- drug-related problems among peo- port is to present estimates about the occurrence of tobacco, alcohol, and ple seeking emergency medical care, other drug use among school-attend- among inmates in correctional facili- 1 Ministry of Health, Panama City, Panama. ties, and among people seeking sub- 2 Ministry of Health, Santiago, Chile. ing youths in Panama, as assessed in 3 The Johns Hopkins University, School of Hygiene the country’s 1996 National Youth stance abuse treatment (1). Additional and Public Health, Department of Mental Hygiene, Survey on Alcohol and Drug Use. need for a national survey also stems Baltimore, Maryland, United States of America. Submit correspondence and reprint requests to: The National Youth Survey is the from the findings of two prior surveys James C. Anthony, Ph.D., Johns Hopkins Univer- first in Panama to examine drug use conducted in Panama. A household sity, School of Hygiene and Public Health, 624 N. Broadway, 8th Floor, Baltimore, Maryland 21205 and suspected determinants of drug survey of drug use conducted in four USA. use using a sample designed to repre- urban areas in Panama in 1991 and a

Rev Panam Salud Publica/Pan Am J Public Health 5(1), 1999 9 school survey in the district of Chan- from a complete list of the schools’ laboration with two of the authors, guinola in 1993 showed the percent of classrooms. Every student attending Luis Caris and James C. Anthony. Lo- lifetime cocaine use to be approxi- the sampled classrooms was recruited gistical and financial support came mately 4.4 in the early 1990s, a percent for participation. The characteristics from CICAD and the Pan American much higher than in several other of the survey respondents are shown Health Organization. Latin American countries in that same in Table 1. To ensure language and conceptual period (2, 3). Assessments were administered as equivalency, the research group trans- This apparent increase in drug use anonymous self-report questionnaires lated the DUSI into Spanish and then in recent years does not seem to be in each class by trained survey staff “back-translated” that document into unique to Panama. According to two members who first introduced the English. The resulting, final Spanish- reports on drug use in Latin America survey and answered questions. The language version of this questionnaire by the Executive Secretariat of the questionnaire was an adapted, Span- instrument has been used extensively Inter-American Drug Abuse Control ish-language version of the Drug Use during the last four years in Central Commission (CICAD) of the Organi- Screening Inventory (DUSI). The DUSI and South America. Furthermore, in zation of American States (OAS), drug was developed to assess alcohol and 1994, a pilot test of the DUSI was con- use appears to be increasing, particu- other drug use, school adjustment, and ducted in Panama with 524 students larly among young persons (4, 5). Re- behavioral and psychological prob- aged 12–19 years; questions the stu- cent increases in cocaine use in Central lems among youths in the United dents found to be unclear or confusing America is a case in point (5). Recently States (9). The DUSI was translated were revised in preparation for the na- documented increases in marijuana and adapted by investigators of the tional survey. use in Chile, as well as the high inci- Ministry of Health in Panama in col- A total of 39 assessors were trained dence of smoking of pasta base (un- for a period of one week before they processed cocaine) in the northern conducted the assessments. Supervi- cities of the country, are additional ex- sors closely monitored the fieldwork amples of the complex drug problem throughout the entire data collection found in Latin America (6–8). TABLE 1. Demographic characteristics of period. For quality control, regional students 12–18 years old participating The results of the national survey supervisors edited completed surveys in National Survey of Alcohol and Other in Panama are intended to guide the Drug Use, Panama, 1996 before forwarding them to the national allocation of resources to facilitate the field office. This work entailed no development, implementation, and n = 6 477 changes in students’ answers, but pro- evaluation of preventive intervention No.a % vided monitoring and rapid detection programs targeted towards the young Characteristic (unweighted) (weighted) of missing data and unclear responses. population of the country. This is an This information then was provided as b important task in view of the increas- Sex feedback to assessors so they would Male 2 788 50.2 ing use of alcohol and other drugs, Female 3 073 49.8 have uniform specifications about how and related health and psychosocial to answer students’ questions. problems. Age Informed consent procedures in- 12 1 399 22.2 13 1 869 29.1 volved notifying parents and allowing 14 1 668 25.8 either students or parents to withdraw MATERIALS AND METHODS 15 1 007 15.4 or to skip questions. Assurances of 16 368 5.3 anonymity were provided. Though The Panama population under study 17 117 1.5 parents might have kept their children 18 49 0.7 was all school-attending youths 12–18 home on the day of the survey in order years old in 1996, except those who Grade level to avoid study participation, to our 1 2 612 40.5 had progressed to postsecondary edu- 2 2 148 33.8 knowledge no parent or student re- cation, that is, to the university level. 3 1 713 25.7 fused participation in the survey. The sample consisted of 6 477 students Missing 4 — Some students skipped sensitive ques- selected via multistage probability School Type tions, as seems to be common for sur- sampling from 154 public and private Public 5 307 83.7 veys of this type. Furthermore, stu- schools, chosen to represent the urban Private 1 170 16.3 dents were told that they could decline and rural areas of the entire country. Location of School to participate at any time without fear- There is no 100% list of youths ages Urban 5 003 78.9 ing retaliation. 12–18 attending school in Panama, Rural 1 474 21.1 To measure the occurrence of drug from which a strictly random sample a Numbers (No.) are unweighted. Percents are weighted use, students were asked the following might be drawn. Hence, schools were and poststratified to reflect the population of school stu- question regarding drug use: “¿Has con- dents in Panama. sampled from a complete list of b Based on the 5 861 students who reported their sex sumido (nombre de la droga) alguna vez en schools, and classrooms were sampled (male versus female). la vida?” [“Have you ever used (name

10 González et al. • Occurrence of tobacco, alcohol, and other drug use among students in Panama of drug)?”], as well as follow-up ques- respective 95-percent confidence inter- reported using marijuana, cocaine hy- tions, such as to assess recency of use. vals (95% CI) were derived by the Tay- drochloride, crack cocaine, pasta base The lifetime experience question was lor series approximation method using (coca paste), hallucinogens, and heroin. asked for tobacco, alcohol, marijuana, Epi Info 6.0 CSAMPLE software (10). Table 3 shows estimates for recent tranquilizers, inhalants, stimulants, co- This method allows for variation in use of tobacco, alcohol, and other caine hydrochloride (cocaine in pow- sampling weights; poststratification drugs. An estimated 33.6% of students der form), crack cocaine (the form that adjustment to such known school pop- used alcohol in the year prior to as- can be smoked), pasta base (coca paste), ulation characteristics as age, sex, re- sessment (95% CI = 30.2 to 37.1), and hallucinogens, and heroin. To measure gion, and public versus private; and past-year tobacco use was 8.7% (95% recent drug use, students were asked survey “design effects” created be- CI = 6.8 to 10.6). The corresponding the following question: “¿Cuántas veces cause students are sampled within estimate for the use of other drugs has consumido (nombre de la droga) en los classrooms and schools rather than was 4.6% (95% CI = 3.6 to 5.5). últimos 12 meses?” [“How may times randomly sampled from a 100% list of have you consumed (name of drug) in all students in Panama. This is a de- the past 12 months?”]. This question is scriptive survey without formal hy- Sex Differences measured on the following 6-point pothesis testing; P values are reported scale: none (1), less than 12 times per only as an aid to interpretation. “Cor- As shown in Table 4, in general, year (2), 1–2 times per month (3), 3–9 rections” for multiple comparisons are male-female differences in drug use times per month (4), 10–20 times per controversial in the biostatistical field were not appreciable, though there month (5), and more than 20 times per (11), and they have not been applied. was a tendency for more males than month (6). The above categories were females to have used licit and illicit recoded as “zero” if the student did not drugs. For example, with respect to al- use in the past year (response category RESULTS cohol, the estimates for use were 45.5% 1) and as “one” if the student had used for males and 37.0% for females. The any time in the past year (response cat- Table 2 shows survey-based esti- corresponding estimates for tobacco egories 2–6). As a result of the low life- mates for lifetime prevalence of use of were 12.2% for males and 9.7% for time prevalence of drug use other than the 11 drugs under investigation. The females. For tranquilizers and stimu- tobacco and alcohol, estimates for re- drug used by the most students is lants, the sex ratio was reversed. For cent drug use were obtained solely for alcohol (41%; 95% CI = 38.3 to 44.2), example, with respect to tranquilizers, tobacco, alcohol, and the category of followed by tobacco (11%; 95% CI = an estimated 7.3% of females had used “other drugs,” which refers to any one 9.3 to 12.6). The most common illicit tranquilizers, versus 6.1% of males. of the following substances: marijuana, drugs are tranquilizers (6.7%; 95% CI = Table 5 shows that male-female dif- tranquilizers, inhalants, stimulants, co- 5.8 to 7.5), inhalants (2.4%; 95% CI = ferences in recent tobacco, alcohol, and caine hydrochloride, crack cocaine, 1.9 to 2.9), and stimulants (1.9%; 95% other drug use also were not apprecia- pasta base, hallucinogens, and heroin. CI = 1.4 to 2.4). As shown in Table 2, ble, though more males (9.5%) than By general convention in drug use fewer than one percent of the students females (7.9%) had used tobacco and epidemiology, “lifetime prevalence” refers to the cumulative occurrence of drug experience up to and including the time of assessment. Its magnitude is conveyed by the size of a proportion TABLE 2. Estimates for lifetime prevalence of tobacco, alcohol, and other drug use among (range: 0–100%). Lifetime prevalence school students, Panama, 1996 (n = 6 477) has the advantage of reflecting the occurrence of drug use in the group as- Estimates of prevalencea sessed during the survey fieldwork. It Drug No. % 95% CI has the disadvantage of not reflecting the experience of potential respondents Tobacco 711 11.0 9.3–2.6 Alcohol 2 631 41.2 38.3–44.2 who are absent or not available for as- Marijuana 56 0.7 0.5–1.0 sessment when the survey is com- Tranquilizers 437 6.7 5.8–7.5 pleted. Thus, as discussed later in this Inhalants 143 2.4 1.9–2.9 paper, to the extent that absenteeism or Stimulants 130 1.9 1.4–2.4 school-leaving is associated with drug Cocaine hydrochloride 42 0.6 0.4–0.9 Crack cocaine 34 0.5 0.3–0.7 use, lifetime prevalence from school Pasta base 34 0.4 0.2–0.6 surveys yields an underestimate of the Hallucinogens 48 0.6 0.4–0.9 actual cumulative occurrence. Heroin 25 0.3 0.2–0.5 All estimates of lifetime prevalence a The numbers of drug users (No.) in the sample (n) are unweighted. Estimates for prevalence proportions (%) are weighted and of recent use of each drug and the and poststratified to reflect the population of school students in Panama.

Rev Panam Salud Publica/Pan Am J Public Health 5(1), 1999 11 TABLE 3. Estimates of recent use of tobacco, alcohol, and other drugs among school past year (see Table 5). On the other students, Panama, 1996 hand, the use of other drugs tended to be greater among those 14–15 years Estimates of past-year usea old (6.5%) than those 12–13 (3.3%) and No. who No. 16–18 (3.3%) (P = 0.05). Drug responded of past users % 95% CI

Tobacco 5 235 468 8.7 6.8–10.6 Alcohol 5 481 1 838 33.6 30.2–37.1 Grade-level Differences Other drugsb 5 098 249 4.6 3.7–5.5 a The number of respondents varies due to missing data for questions on alcohol, tobacco, and other drug use in the past year. The occurrence of drug use ad- The number of drug users in the sample is unweighted. Estimates for prevalence proportions (%) are weighted and poststrat- vanced in magnitude more regularly ified to reflect the population of school students in Panama. with respect to grade level than in re- b Category of “other drugs” includes marijuana, tranquilizers, inhalants, stimulants, cocaine hydrochloride, crack cocaine, pasta base, hallucinogens, and heroin. lation to age (see Table 4). For example, an estimated 29% of students in secondary grade level 1 had consumed alcohol at least once in their lifetimes. The corresponding estimates for stu- more males (36.7%) than females tobacco, alcohol, and several other il- dents in grade levels 2 and 3 were (30.5%) had used alcohol in the year licit drugs (e.g., marijuana and cocaine 42.3% and 59.1%, respectively. Simi- prior to the interview. The use of other hydrochloride). For example, while larly, tobacco had been used by 7.0%, drugs was greater for females (5.5%) 32.9% of those 12 or 13 years old had 10.7%, and 17.4% of students in sec- than for males (3.8%), but this was due consumed alcoholic beverages, the es- ondary grade levels 1, 2, and 3, respec- to the greater percentage of females timate for those 14–15 years old was tively. who used tranquilizers, as indicated 40.0% and for those 16–18 years old it Recent tobacco, alcohol, and other above. was 55.5% (P = 0.01). This consistency drug use also increased with grade of age-related differences was not seen level. For example, the proportion of for tranquilizers, inhalants, stimulants, students who had used alcohol in Age Differences pasta base, or hallucinogens. the past year in grade level 1 was Similarly, a greater proportion of 21.5%. That rose to 34.9% and 50.4% in A greater proportion of older stu- older students than younger students grade levels 2 and 3, respectively (see dents than younger students had used had used tobacco and alcohol in the Table 5).

TABLE 4. Estimated lifetime prevalence of tobacco, alcohol, and other drug use by demographic characteristics (weighted percents), Panama, 1996a

Tobacco Alcohol Marijuana Tranquilizers Inhalants Total sample size Characteristic (unweighted) No. % No. % No. % No. % No. %

Sex Male 2 788 350 12.2 1 267 45.5 32 1.0 169 6.1 74 3.0 Female 3 073 300 9.7 1 150 37.0 16 0.6 226 7.3 58 1.9 Age 12–13 3 268 235 7.5 1 064 32.9 17 0.4 173 5.1 59 2.0 14–15 2 675 394 14.5 1 286 40.0 26 1.0 221 8.4 71 3.0 16–18 534 82 15.6 281 55.5 13 2.1 43 7.7 13 2.5 Secondary grade level 1 2 612 174 7.0 735 29.0 11 0.4 137 5.3 42 1.8 2 2 148 232 10.7 885 42.3 20 0.6 141 6.4 47 2.5 3 1 713 305 17.4 1 009 59.1 25 1.5 158 9.2 54 3.3 School type Public 5 307 466 8.8 1 957 38.4 45 0.7 327 6.2 105 2.2 Private 1 170 245 21.8 674 55.7 11 0.9 110 8.9 38 3.4 Location of school Urban 5 003 615 12.4 2 215 44.2 43 0.8 354 6.9 118 2.4 Rural 1 474 96 5.6 416 30.1 13 0.7 83 5.9 25 2.7

12 González et al. • Occurrence of tobacco, alcohol, and other drug use among students in Panama TABLE 4. (continued)

Cocaine Stimulants hydrochloride Crack cocaine Pasta base Hallucinogens Heroin Total sample size Characteristic (unweighted) No. % No. % No. % No. % No. % No. %

Sex Male 2 788 54 1.8 25 0.8 18 0.6 20 0.6 28 0.8 14 0.4 Female 3 073 56 1.9 12 0.4 12 0.4 9 0.3 13 0.4 7 0.3 Age 12–13 3 268 53 1.5 15 0.4 12 0.4 14 0.4 26 0.7 8 0.2 14–15 2 675 68 2.4 18 0.8 16 0.6 18 0.5 18 0.6 12 0.5 16–18 534 9 2.0 9 1.4 6 1.0 2 0.4 4 0.6 5 0.8 Secondary grade level 1 2 612 39 1.3 12 0.5 14 0.5 10 0.3 24 0.8 7 0.2 2 2 148 42 1.8 15 0.5 6 0.2 15 0.5 14 0.5 10 0.3 3 1 713 49 2.9 15 1.0 14 0.9 9 0.5 10 0.5 8 0.6 School type Public 5 307 105 1.9 37 0.6 33 0.6 29 0.4 39 0.6 24 0.4 Private 1 170 25 1.5 5 0.4 1 — 5 0.5 9 0.7 1 — Location of school Urban 5 003 92 1.6 30 0.6 27 0.5 23 0.4 38 0.6 18 0.3 Rural 1 474 38 2.7 12 0.7 7 0.4 11 0.5 10 0.5 7 0.3 a Numbers of drug users (No.) are unweighted. Percents are weighted and poststratified to reflect the population of school students in Panama.

Public versus Private Schools observed for tobacco (6.4% for public DISCUSSION versus 19.9% for private school stu- As shown in Table 4, alcohol use was dents, P = 0.01) and alcohol (31.1% for The evidence from this study, based much more likely to have occurred public versus 46.2% for private school upon the first-ever nationally repre- among private school students (57.7%) students, P = 0.01). sentative sample survey in Panama of than among public school students students 12–18 years old, is consistent (38.4%) (P = 0.01). The same pattern with evidence from many other places. was observed for the use of tobacco, in- Urban versus Rural Schools Specifically, the evidence tends to halants, and tranquilizers. More pri- show more males than females using vate school students reported tobacco Experience with tobacco and alcohol alcohol, tobacco, and inhalants. Stu- use than public school students (21.8% was substantially higher in urban dents in higher grade levels and those versus 8.8%, P = 0.01), and private areas than in rural areas (see Table 4). in urban schools were more likely to school students were also more likely The most striking differences involved have started using drugs. Public to have used inhalants (3.4% versus tobacco and alcohol: an estimated school students were less likely than 2.2%, P = 0.05). Finally, a greater pro- 12.4% of students in urban areas had private school students to have started portion of students in private school smoked tobacco versus an estimated drug use, a finding consistent with (8.9%) than in public schools (6.2%) had 5.6% in rural areas. The corresponding data from similar studies in other used tranquilizers (P = 0.01). Public- estimates for alcohol use were 44.2% Latin American countries (12–15). private school differences on the use of for urban students and 30.1% for rural There is a consistent pattern of greater other drugs (e.g., heroin and stimu- students. With respect to the other alcohol, tobacco, and illicit drug use lants) were not appreciable (P > 0.05). drugs, no urban-rural differences were among males than among females. Table 5 shows that recent use of to- striking. However, there might be important bacco, alcohol, and other drugs was Similarly, a greater percentage of gender differences from country to more common among private school students attending urban than rural country. For instance, in Panama more students. The difference in the use of schools had recently used tobacco females than males use tranquilizers, other drugs between public and pri- (9.8% versus 4.1%, P = 0.01) and alco- but in Colombia the difference is not vate school students was statistically hol (35.3% versus 26.5%, P = 0.01) (see statistically significant. In Chile more significant by conventional standards Table 5). Urban-rural differences in females than males use tobacco, a pat- (4.2% versus 6.8%, respectively, P = past-year use of other drugs were neg- tern also true for some regions of 0.01), though it was smaller than that ligible (P > 0.05). Brazil (13). Our findings also show a

Rev Panam Salud Publica/Pan Am J Public Health 5(1), 1999 13 TABLE 5. Estimates for past-year use of tobacco, alcohol, and other drugs by demographic drug use (20). These investigators characteristics, Panama, 1996 found that respondents using the audio-CASI method were three times Total no. of Tobacco Alcohol Other drugsb respondentsa more likely to report injection drug Characteristic (unweighted) No. % No. % No. % use than respondents using the standard self-administered questionnaire; Sex however, it is not clear that greater re- Male 2 228 227 9.5 887 36.7 85 3.8 porting is equivalent to more accurate Female 2 476 202 7.9 819 30.5 139 5.5 reporting. The results of this and other Age studies demonstrate the need for 12–13 2 679 153 6.0 702 25.1 91 3.3 14–15 2 191 270 11.5 937 41.0 144 6.5 more research on methods such as hair 16–18 365 45 12.7 199 51.7 14 3.3 analysis or computer-assisted technology that might increase the accuracy of Secondary grade level 1 2 101 106 5.3 470 21.5 71 3.2 survey estimates. However, we are 2 1 735 155 8.9 616 34.9 79 5.1 most likely many years from use of 3 1 395 207 13.8 751 50.4 98 6.4 these expensive (and sometimes inva- School type sive) methods in school surveys of stu- Public 4 194 276 6.4 1 328 31.1 174 4.2 dent drug use with sample sizes in the Private 1 041 192 19.9 510 46.2 75 6.8 thousands, at least in the poorer coun- Location of school tries of the Region of the Americas, Urban 4 210 412 9.8 1 561 35.3 201 4.6 where research and surveillance re- Rural 1 025 56 4.1 277 26.5 48 4.9 sources are in short supply. In this re- a The number of respondents varies for the categories of alcohol use and “other drugs” use due to missing values. The num- gard, we note the utility of the “Moni- ber of drug users in each category (No.) is unweighted. Estimates for prevalence proportions (%) are weighted and post- toring the Future” school surveys in the stratified to reflect the population of school students in Panama. b The category of “other drugs” includes marijuana, tranquilizers, inhalants, stimulants, cocaine hydrochloride, crack cocaine, United States (26) and similar surveys pasta base, hallucinogens, and heroin. in Canada, where there has been complete reliance upon self-report methods. A second limitation is that this survey provides estimates of drug use for greater occurrence of alcohol, tobacco, to be impractical in school surveys school-attending youths only. Given and other drug use among private (16). A review of recent research on the the high dropout rates among interme- school students as compared to those validity of self-reported drug use and diate school students in Panama (21), in public schools. This finding also is the methodological advances under- it is possible that drug use among consistent with findings from similar taken to reduce error in estimates of Panama’s total population of youths is studies conducted in other Latin drug use concluded that the use of underestimated when surveys are American countries (13, 14). The au- self-report is less valid for more stig- based on school sampling frames. In thors have hypothesized that students matized substances, such as cocaine, Panama, there were an estimated in private schools are of higher socio- and for more recent drug use (17–19). 137 054 students in public and private economic status and may have more These and other reports suggest that schools attending the grade levels purchasing power and thus may be one viable alternative to assessment of under study. Though not part of the more able to afford alcohol, tobacco drug use might be the use of hair population under study, approxi- products, and illegal drugs. The rea- analysis, saliva cotinine assessments, mately 9% of the school age popula- sons for these differences in student or other bioassay validation methods tion in Panama does not attend school drug use observed in Latin America (17). Another strategy suggested to (22). Future studies that investigate the remain understudied; future studies minimize response bias is the use of extent of drug use among school should investigate them more deeply. audio computer-assisted self-interview- dropouts are needed to obtain a more Before a more detailed discussion, ing (audio-CASI). Audio-CASI tech- accurate assessment of youth drug in- findings from this report can be viewed nology involves using a computer to volvement in Panama. Third, a num- in light of several limitations. First, the administer a survey in an audio for- ber of students skipped the question data are based on retrospective self- mat and record responses without the on recent drug use that asks about reports of drug use. The use of self- involvement of the assessor. In a re- frequency of drug use in the past 12 reports may have introduced errors in cent study of drug use among adoles- months for each of the 11 drugs listed recall and perhaps a socially desirable cent males in the United States of in Table 2. This problem raises ques- response tendency. The use of addi- America, researchers utilized the audio- tions about the validity and reliability tional measurement methods (e.g., bio- CASI method and a standard self- of the measure for recent drug use. To assays) might assist in establishing the administered questionnaire to obtain examine this problem more carefully, validity of the responses, but they tend estimates of the prevalence of injecting future surveys should include a sepa-

14 González et al. • Occurrence of tobacco, alcohol, and other drug use among students in Panama rate question for each drug and ask some degree of underestimation in recommend that careful outcome and students whether they have used any self-report surveys. process evaluation be conducted when of them in the past 12 months. This One of the most interesting results implementing any prevention program. question can then be followed by ques- from this survey involves tranquilizer In conclusion, the observed occur- tions on frequency of use for each of use, which is seen primarily among fe- rence of alcohol and tobacco use, even the drugs in the past year. males. Students in Panama are much at the lower school grade levels stud- Notwithstanding these limitations, more likely to have started using tran- ied in the survey, makes it important these findings have implications for quilizers than those at comparable to enhance primary prevention activi- the Panamanian drug prevention and grade levels in the United States (26). ties in Panama. Curtailing the inci- policy efforts. Primarily, these efforts Anecdotal reports suggest a practice in dence of licit and illicit drugs today is concentrate on preventing initiation of Panama, also somewhat common in important in order to prevent an in- alcohol and tobacco use as a means to other countries in Latin America, in crease in the number of youths who avert health and social problems com- which parents give their children tran- are engaged in high-risk behaviors, monly associated with the use of these quilizers when the youths are experi- such as unsafe sex. These behaviors or other substances. However, it is of encing stressful circumstances. Future may lead to other detrimental and seri- concern that the apparent risk of de- studies are needed to clarify this situa- ous consequences, including infection veloping alcohol or other drug prob- tion and perhaps to impose greater re- with the human immunodeficiency lems is increased for persons who start strictions upon the availability of these virus or other sexually transmitted dis- using these substances at an earlier age tranquilizers. eases. Findings from the present sur- (e.g., before the age of 15 versus 18–21) The observed prevalence estimates vey present the leadership in Panama (23, 24). Moreover, at least in some for illicit drug use create a unique op- with an opportunity to target preven- samples, use of tobacco and alcohol portunity for leadership to mobilize tion efforts among school-aged youths, runs in advance of use of such other and concentrate efforts on primary before drug use and other public drugs as marijuana (25). prevention activities in Panama. These health problems related to drug con- It is important to note that fewer activities might help to maintain the sumption become more prevalent. than one percent of the students par- observed low occurrence rate of illicit ticipating in the survey indicated they drug use. Or, if the true values are had ever used marijuana, cocaine hy- larger than those reported here, then drochloride, crack cocaine, heroin, or the prevention efforts can help to sta- Acknowledgements. This project hallucinogens. These findings are in bilize this situation, prior to reductions was supported by the United Nations contrast to trends observed for the in drug use prevalence. The array of Educational, Scientific, and Cultural Panamanian population groups in the available prevention interventions is Organization (UNESCO), the Inter- surveillance data based on emer- broad. They include school- and com- American Drug-Abuse Control Com- gency room and incarcerated samples. munity-based prevention programs, mission of the Organization of Ameri- Whether these apparent differences as well as family- or individually- can States (CICAD/OAS), the Pan between population segments are due oriented programs that can be used to American Health Organization to variation in the sample composi- target such different risk factors as (PAHO), the “Coalición de Panamá tions or validity of self-report methods association with deviant peers, family por una Comunidad Libre de Drogas, remains on the agenda for future re- inattention, and drug availability Despacho de la Primera Dama de search. In addition, the 1996 evidence (27–30). However, we suggest caution Panamá,” and training grant T32- provides a baseline against which to in transplanting prevention programs DA07292 from the U.S. National Insti- gauge future trends, even if there is from other countries to Panama and tute on Drug Abuse.

REFERENCES

1. Ministerio de Salud de Panamá. Vigilancia epi- y Desarrollo de la Agencia de los Estados americano de datos uniformes sobre consumo demiológica del uso de sustancias psico-activas en Unidos para el Desarrollo Internacional; 1991. de drogas, volumen I: América del Sur. Centroamérica, República Dominicana y Panamá. 3. Cruz Blanca Panameña. Resultados de la pri- CICAD/OAS; 1997. Organización de Estados Americanos, Organi- mera encuesta de consumo de alcohol y otras 5. Executive Secretariat of the Inter-American zación Panamericana de la Salud, y MSP; 1996. drogas realizada en la provincia de Bocas del Drug Abuse Control Commission of the Orga- 2. Centro de Estudios Latinoamericanos y Pro- Toro [unpublished report]. Cruz Blanca Pana- nization of American States. Sistema interame- yecto Narcotics Awareness and Education. meña y el Centro de Estudios de Prevención ricano de datos uniformes sobre consumo de Resultados del primer estudio epidemioló- de la Universidad de Colorado; 1993. drogas, volumen III: América Central, Panamá gico sobre prevalencia y actitudes del uso de 4. Executive Secretariat of the Inter-American y República Dominicana. CICAD/OAS; 1997. drogas en el área urbana de Panamá [unpub- Drug Abuse Control Commission of the Or- 6. Chile, Ministerio del Interior, Consejo Na- lished report]. Departamento de Investigación ganization of American States. Sistema inter- cional para el Control de Estupefacientes.

Rev Panam Salud Publica/Pan Am J Public Health 5(1), 1999 15 Consumo de drogas en Chile: síntesis de los en escolares año 1997. Chile. Ministerio de Edu- 24. Grant BF, Dawson DA. Age of drinking onset principales estudios y datos estadísticos. cación, Ministerio de Salud, Consejo Nacional predicts future alcohol abuse and depen- CONACE; 1996. para el Control de Estupefacientes y Fun- dence. J Subst Abuse 1997;9:103–110. 7. Chile, Ministerio del Interior, Consejo Na- dación Paz Ciudadana; 1998. 25. Kandel DB, Yamaguchi K, Chen K. Stages of cional para el Control de Estupefacientes. Se- 16. Anthony JC. Do I do what I say?: A perspec- progression in drug involvement from ado- gundo informe annual sobre la situación de tive on self-report methods in drug depen- lescence to adulthood: Further evidence for drogas en Chile. CONACE; 1997. dence epidemiology. In: Stone A, Turkkan J, the gateway theory. J Stud Alcohol 1992;53: 8. Caris, L. Tabaco, alcohol y drogas. Ministerio Bachrach CA, Cain V, Jobe JB, Kurtzman H, 447–457. de Salud, Chile. 1998. Available from: URL: eds. Science of self-report: Implications for re- 26. Johnston LD, O’Malley PM. National survey http://www.minsal.cl search and practice. Hillsdale, NJ: Lawrence results on drug use: The Monitoring the Future 9. Tarter RE, Hegedus AM. The Drug Use Earlbaum Associates, Inc. In Press. Study. Available from: URL:http://www.isr. Screening Inventory: Its applications in the 17. Harrison L, Hughes A. Introduction—The va- umich.edu/src/mtf/ evaluation and treatment of alcohol and other lidity of self-reported drug use: Improving 27. Hawkins J, Catalano R, Miller J. Risk and pro- drug abuse. Alcohol Health and Research World the accuracy of survey estimates. In: Harrison tective factors for alcohol and other drug 1991;15(1):65–75. L, Hughes A, eds. The validity of self-reported problems in adolescence and early adulthood: 10. Dean AG, Dean JA, Coulombier D, Brendel drug use: Improving the accuracy of survey esti- Implications for substance abuse prevention. KA, Smith DC, Burton AH, et al. Epi Info, Ver- mates. U.S. Department of Health and Human Psychol Bull 1992;112(1):64–105. sion 6: A word processing, database, and statistics Services (NIH Pub. No, 97-4147) (National In- 28. Grover PL. Preventing substance abuse among program for epidemiology on microcomputers. At- stitute on Drug Abuse Research Monograph children and adolescents: Family centered ap- lanta, Georgia: U.S. Centers for Disease Con- No. 167); 1997:1–16. proaches. U.S. Department of Health and Hu- trol and Prevention; 1994. 18. Harrison L. The validity of self-reported data man Services, Substance Abuse and Mental 11. Rothman KJ. Modern epidemiology. Boston, on drug use. Journal of Drug Issues 1995; Health Administration, Center for Substance MA: Little, Brown, and Company; 1986. 25(1):91–111. Abuse Prevention 1998. (U.S. DHHS Publica- 12. Torres Y, Maya Mejía JM, Murelle L. Estudio 19. United States of America, General Account- tion No. 3223-1998). nacional de consumo de sustancias psicoactivas y ing Office. Drug use measurement: Strengths, 29. Zucker R, Boyd G, Howard J., eds. The devel- factores asociados. Estudiantes de educación se- limitations, and recommendations for improve- opment of alcohol problems: Exploring the bio- cundaria y media. Colombia. Colombia: Vicemi- ment [unpublished report]. Report to the psychosocial matrix of risk. U.S. Department of nisterio de la Juventud; 1996. Chairman, Committee on Government Op- Health and Human Services; (Research Mon- 13. Galduróz JCF, Regina Noto A, Carlini EA. erations, U.S. House of Representatives (GAO/ ograph No. 26.) (National Institutes of Health Tendencias do uso de drogas no Brasil: Sín- PEMD-93); 1993. Publication No. 94-3495). tese dos resultados obtidos sobre o uso de 20. Turner CF, Ku L, Rogers SM, Lindberg LD, 30. Botvin GJ, Schinke S, Orlandi MO, eds. Drug drogas entre estudantes do 1o e 2o graus em 10 Pleck JH, Sonenstein FL. Adolescent sexual abuse prevention with multiethnic youth. Thou- capitais Brasileiras (1987–1989–1993–1997). behavior, drug use, and violence: Increased sand Oaks, CA: Sage Publications; 1995. Universidade Federal de Sao Paulo, Escola reporting with computer survey technology. Paulista de Medicina, Centro Brasileiro de In- Science 1998;280:867–871. formaçoes sobre Drogas Psicotrópicas; 1997. 21. Reddy MA. Statistical abstract of the world. 14. Alcaraz del Castillo F, Soliz Vázquez RM, New York: Gale Research Inc; 1994. Zuaso Yujra J. Consumo de alcohol, tabaco, 22. Pan American Health Organization. Health in cocaína y otras drogas en Bolivia. Bolivia: the Americas, volume II. Washington, D.C.: Centro Latinoamericano de Investigación PAHO; 1998 Científica; 1998. 23. Anthony JC, Petronis, KR. Early-onset drug 15. Caris, L. Informe preliminar del segundo estudio use and risk of later drug problems. Drug Al- Manuscript received on 23 February 1998. Revised ver- nacional del consumo de alcohol, tabaco y drogas cohol Depend 1995;40(1):9–15. sion accepted 11 December 1998.

RESUMEN Este informe presenta las primeras pruebas epidemiológicas del consumo de tabaco, alcohol y otras drogas en alumnos de escuelas panameñas, usando datos obtenidos mediante una encuesta estudiantil completada en 1996. En particular hemos exami- Uso estimado de tabaco, nado las diferencias por sexo, edad, grado escolar, tipo de escuela y residencia urbana alcohol y otras drogas en o rural en el uso de tabaco, alcohol y otras drogas. Se estimaron la prevalencia vitali- estudiantes panameños de cia y el consumo previo de estos productos a partir de una encuesta nacional de la juventud efectuada en 1996 (n = 6 477). Debido al diseño muestral multietápico aplicado 12 a 18 años: resultados de la en la encuesta, todas las estimaciones y sus respectivos errores estándar se derivaron Encuesta Nacional de Panamá por el método de Taylor de series aproximadas aplicando el programa Epi Info 6.0 sobre el Uso de Alcohol y CSAMPLE. Drogas en la Juventud En general, los varones, los alumnos de mayor edad y los alumnos de grados su- periores fueron los que mayormente habían consumido drogas lícitas e ilícitas, aunque fue pequeña la diferencia entre varones y hembras. Las difrencias entre alumnos de escuelas públicas y colegios privados y las tendencias observadas en habi- tantes de zonas urbanas frente a los de zonas rurales variaron según el tipo de droga. Los resultados de este estudio se examinan a la luz de la epidemología y prevención del consumo de drogas en Panamá. Partiendo de estos datos, esperamos poder pro- veer al Gobierno de Panamá información que le sea útil para planificar programas de prevención dirigidos a alumnos de escuelas panameñas.

16 González et al. • Occurrence of tobacco, alcohol, and other drug use among students in Panama