KEY wc?Ems:
Ackolescent Ctontcaceptiveuse Pregnancy prevention
cocs*Frolhy adolescents h,as al deb& md mearch over the research has ir.vesLigated factors adoption of a contraceptive devoted to underon&tent and effective use of a birth coutrol method (I-3). Many sexually active adoiescmts do traception effectively or consistently.
Address feprint requests to: May km Balmone, D.S.W., School of Swid Work JH-30, University of Washington, Seattle, WA1 98295. fvbmscript accepid juk -, *_4?9.
Table 1 sumnarizes the endings of ~~e~o~s research investigating the con& contraceptive use by tee of birth control vary, methods as well as the these results suggest th gories of variables in predict ceptive use at first con nancy history as we aspimtkns and views of self successful birth control use. adolescenYs int9-p sandp initial co relate tracept+e
users regardiq
0 Sociely for Adolescent Medicine, 1989
PdIidwd by Ekevier science Publishing Co., lnc ,655 Avenue of the Americas, New Yak, NY lC010
correlates of contimed 527
JOURNAL OF ADOLESCENT
BALAsom
528
be renewed
Inconsistent Birth Control Use Among Adolescents
Table 1.
Comlates of discontinuation
Deanogaphicat &cialchWacte&tic3 Yaw ‘~ppat first birth conml use &ck or hispank No method at first intercourse Prkious abortion Individual differences Lower educational aspirations Low ievels of personal health Low frustraiion tolerance Rkperience problems/side effects in using method Feelings of helplessness Interpersonal relationships Lacks steady committed sexual relationship/ infquent sex Wents have no knowledge of contraceptive
Reference no. (8,17-19) (8) (20) (21) (83) w (22) (l&22,23) &)
after 3 months. The latter follow-up ivhen the subjects returned to planning clinic. AlI interviews with teenagers at the initial visit were conducted by the researcher who toll demographic, reproductive history, and psychosocial data. Demographic variables were age, education, race,, and hou productive heahh history a&e, .age of first
r perception of oral contrace
(8,10,19,21) (18)
U8e
Provider/clinic chracteristics D&&i&d with method/visit Did nut make/pay for own appointment
LTH CARE VOI. IQ, No. 6
(8,Iq,l8,24,25) <18,19,2i)
use of birth controL Two research questions were addressed: a) What factors (demographic, cognitive, and psychosociaL) distingui& adok&ents who continue using an oral contraceptive from those who do not? b) What specific problemi: do teenagers encounter in using an oral contrace@e?
Methodi This study uses a ve research design with two major data co ints. The research procedures were approved by the Human Subjects Review co s at both the University of Gbf&nia, &rkeley and the participating family planning agency. Adolescents who planned to use an oral contraceptive were interviewed at their initial family planning clinic visit. Tl?&e clinics run by a private family planning agencJ” in Oakland, California serving clients from a satiety of racial groups and socioeconomic levels were used as data collection sites. A sample of 76 adolescents (17 years or younger) who attended randomly selected clinic sessions during the summer of 1986 and received an oral contracept& ,p&d in the study. Au of asked to be part of the study Oral contraceptive users were chosen as the study population because approximately 70% of female teenagers who report using birth control use the piu (6), and an oral contraceptive pr&cr$&n must
contraceptive use and pregnancy. Th and benefit questions came from the and Bitte (26) and Luker (27), who ge prehensive yet parsimonious list of
an oral contraceptive. The subject responded to problem and benefit using a four-point Likert ranging from 1 (no problem) to 4 (major ptoblem)
effect of separate items. Scale reliability was measured using Cronbach’s alpha (28) (a measure of internal consistexy). Reliabfities were 72, 61, .Tp and .86 for the four summary scales of costs of use, benefits of pill use, costs of pregnancy, benefits of pregnanq, respectively. A number of interview items collected information related to the accuracy of edge. One question asked subject likelihood of pregnancy without traceptive (scale 1 = not likely; 4 = very KnowMge about the effectiveness of the various birth control methods was assessed by asking the subject to esGmate ‘he number of teenagers, out of 100, who wovlld get pregnant if they were: not using a method of birth control, using the pill, using a barrier method, or using withdrawal. Each subject was also asked to rate her own risk of pregnancy as compared with the risk faced by other adolescents. This question required the ger to rate her susceptibility on a five-print Bike tale (1 = much less likely to get pregnant than other teenagers, 5 = much more likely to get pregnant).
November 1989
SK OF CONT
0N
Tot4
Characteristic
sample n (%) 15.6
BkXk
Hispanic
e were recon
39 (51) 30 (39) 2 (3) 5 (7)
Dkontinued
use n (%) IS.8 8 6 1 1
(so) 438) (6) (6)
5
Contiue
use ?I (%) 15.5 26 11 1 4
(62) (26) (1) (10)
12.1
12.1
12.1
56 (74) 20 (26)
12 (75) 4 (25)
34 (f-w 8 (19)
MenstrraP periods
Contraceptive history Ever had sex 7 (9) i%uc 69 (91) Yes Age at 1st intercourse (yr) 14.2 Wait before coming to 1.4 clink (yr) Ever used birth control method
1 (6) 25 (94) 14.5
26 (34) 7 (9)
8 (w) 7 (44) 1 (6)
25 (@) 12 (29) 5 (11)
58 (77) 11 (14) 7 (9)
13 (81) 2 (131 1 (6)
30 (71) 7 (18) 5 (11)
57 (75) 12 (16) Yes 7 (9) NA General & gynecologic health history 1.3 Health prohlems
13 (81) 2 (13) 1 (6)
34J (71) 7 (18) 5 (11)
NO Yes
NA Pregnancy history Ever pregnant NO
Yes NA Fears pregnancy now
43
6%
NO
1.4
1.4
Cyn problems
its
Results
were made using
x2 and t-test
None Xad a ST?2 Preview P4P smear None c year ago > year ago Clinical report on visit Exam result Normal Abnormal Lab tests Normal Yeast infection SD i?&Ssing
Referral” No Yes *Xz=~.~;df=1;p=0.08.
75 (99) 1 (I)
16 (100) Q (0)
4: (98) 1 (2’
55 (72) 18 (24) 3 (4)
10 (63) 4 (25) 2 (12)
31 (74) 10 (24) 1 (2)
68 (89) 8 (11)
14 (f37\ 2 (12)
58 (78) 2 (3)
11 (79) 0 (0) 3 (23) 2-
14
[Ic+
2-
70 (92 6 (8;
12 (75) 4 (25)
530
BALASSONE
another method of birth control. Abnost 20% of the sample, although asymptomatic, had a sexmthy transmitted &ease (STQ) diagnosed at the initial clinic visit. Just over half (PJ = 42) of the sample returned to the family planning clinic to renew their oral contraceptive prescription. These adolescents apparently continuedto use the piiiand were categ~rdd as c0mismt oral contraceptive users. Forty-five percent (n = 34) did not return to the clinic. All of the teenagers who were recontacted after failing to return for theirblood pressure check(n = 16) reported discontinuinguse of the pill. Categorizationof all the adolescents who failed to return to the clinicas inconsistentoral contraceptive users was not possible.Subgroup analyses comparing subjects contacted by phone to those who were not recontacted identifieddifferencesin variables related to background, medical history, and psychosocial characteristics. In addition, 11 of the 18 who were not recc&acted had been visitingOakland for the summer and had returnedhome. Those teenagers not recontacted were probably a mix of consistent and inconsistent oral contraceptive users. The bivariate analyses reported in this paper compa co&stent users (adolescents who returned to the clinic) to inconsistent users (adolescents who did not return to the clinic, were interviewed by phone and whose discontinuation of pill use was confirmed). Table 2 summarizes, for the entire sample, de mographic and reproductive history characteristics as well as items related to the adolescent’s mitial clinic visit and also compares conZ;&ent and inconsistent contraceptive users. Teenagers who continued using the pill were not significantly different from those who discontinued use on any of the demographic or reproductive health history variables. The two groups did differ on one item at the initial clinic visit. The group of inconsistent contraceptive users were more likely to need additionac tests (blood tests or pap smear) before receiving more than one package of pills. Table 3 compares selected psychosocial variables for the adolescents who continued to use a prescribed contraceptive and those who discontinued use. The differences between mean scores on the s.mlImary scales related to the costs of pill use, hene& Of pfi USe, CO!& Of yrq$xl;licy, and benefitsof Pregnancy for those adolescents who conthd to use the pill and those who did ~turnr were not statisticalty sigmficant. Teenagers who discontinued using the oral contraceptive were more likely to
Vol. IO, No. 6
JOURNAL OF ADOLESCENT
Table 3. !Mected Psychosocial and Cognitive
Characteristics: Total Sample and Comparison of Adolescents Continuing and Discontinuing Use 0fthePiU Total sample Mean (SD) cost of pill use (scale l-4) Health costs (scale l-4) Pill use benefits (scale l-4) Preglmncy costs (scale l-4) Pregnancy benefits (scale l-4) personal susceptibility to PWWY” (scale l-5) problem sohing’ (no. of options)
Discontinued IvleGse(SD)
Continued Mea?&
1.7
(.5)
1.8
(-5)
1.7
(.5)
2.1
i.7)
2.4
(.7)
1.8
(.6)
3.5
(.6)
3.4
(.6)
3.6
(.5)
2.9
(.7)
2.9
(.6)
2.8
(.7)
1.8
(I)
2.0
(.9)
1.8
(.7)
2.7
(1.0)
2.5
(1.0)
2.9
(.8)
2.8
(.7)
2.5
(-7)
2.9
t.7)
;: 1 2.29; d_f= 56; p = om. - -1.74; af = 56;p = 0.09. ‘t = -1.82; df = 56; p = 0.07.
report a concern problems.
about
potential
health-related
adolescents interviewed a the risk of pregnancy a various birth control methods. Most subjects underestimated the likelihood of pregnancy when not using a birth control method. They estimated the number of resuhing when no method of birth to be le,qs than the actual number e dian sample estimate was 75 pregnanaes nonusers, whereas the actual number of pregnancies expected is 90 (31). Subjects also misjudged the number of pregnancies occurring with the use of a variety of birth control methods. They underestimatedthe effectivenessof oral contraceptivesand barriermethods. Although those teenagers who were consistent oral contraceptive users tended to give more accurate estimates of the effectiveness of birth control and likelihood cy, the differences were not statistically As shown in se adolescentswho to return for follow-q2raterBtheir ~3 risk for pregnancy as iess than that of their peers. Sixteenpercent of the teenagers thought th2y were more an other teenagers to get pregnant, 47% as likely, and 37% less likely. When the latter group were asked for the reasons for this belief, 54% said
532
BALASSONE
how consistent users contend with these problems. momtion on the behaviors of teenagers in this mple may offer some direction for the design of interventions to enhance contraceptive continuation. caution in generalizing the results reported here is warranted due to the small sample size resulting from the loss of cases at fellow-up. Although removing cases where contraceptive use status at follow-up is unknown provides a better criterion measure (i.e., these analyses actually compare continuers and discontinuers), the results should be viewed as suggestive. Replication of this study with an additional sample of adolescents is needed. Characteristics that tend to distinguish consistent pill users from inconsistent users iu this study fell @to two categories (Table 1): individual differences and provider/clinic characteristics. The two groups differed in their view of the health COSBrelated to pill use, personal susceptibility to pregnancy, and problem-solving ability. Adolescents who saw higher health-related costs, held beliefs of decreased personal pregnancy susceptibility, and were less able to identify pregnancy outcome options were less likely to continue contraceptive use. Assessment of these cognitive factors by practitioners may help identify adolescents who are less likely to continue to use the pill. The questions from the research interview are currently being adapted for use by practitioners in a clinic setting. A brief set of yuesh=ts to assess the teenager’s decision to use an oral contraceptive, perception of her susceptibility to pregnancy, view of costs related to pill use, and problemsolving skill may be useful to screen adolescents for a high risk of contraceptive discontinuation. The possible connection between these cognitive fack,rs and actualsor&aceptive behavior suggests that intervention during the initial clinic visit may enhance contraceptive continuation. For example, inconsistent users who see more problems associated with pill Use may have a lower tolerance for side-effects that occur. These teenagers appear to have a reduced appreciation of their own probability of getting pregnant, which also may lesd to less diligence in using contraception. More direct education r%arding the occurren.ce and handling of oral contrawptive side-effects is, one intervention that could be pvided to those adolescents who may be more likely to discontinue contraceptive use. J’his study also suggests that the typical medical protocol in family planning clinics requiring referral for additional tests may interfere with s0m.e keenae’ COntinud use of birth control. Adolescents
JOURNALOF ADOLESCENTHEXLIT- CARE Vo9. IO, No. 6
who were given only one package of to r&ICYa&BY ara ~dditionai test before pi renewal were less likely to continue Extra time devoted to ex+ining the type of test, the specific how-tos of obtain&g the test, and the importance of the results might result in more consistent contraceptive agers may also help e p Adolescents in thi se did better in riaceptiv y wese iiiiore SUCCQS method. to take the pill each day by associating pill taking with another regular activity. Inconsistent users tended to have others remind them or to leave the e tht4r pills in view, and hence, often forgot znagepill. The two groups also differed in t ment of side+ffects. Consistent users were more likely to call the &nic for advise, while inconsistent users tended to just stop taking the pill. Reassmance by staff may have played a role in continued oral contraceptive use because the teenager c.3 vided with accurate information about. problem and its consequences. These f gest that increased teaching of titioners to associate their pilldaily activity and to call the clinic should a side-effect ante contraceptive con Iinuation. as looked only at adolescents who use an oral contraceptive. Although the pill remains the method most commonly used by teenagers who use contraception <32), there is a need for research concerned with other birth control met The increasing incidence of sexually tmnsmitt eases sts a need to better among adolescents (33-35) s understand factors that facilitate both ir&iai and continued use of condoms by teenagers.
References 1. HofmanAD.Contraceptionin adolescence:a review of psy-
chosocialaspects. BullWorld Health Org 1984;52:151-62. 2, MorrisonDM. Adolescentcontraceptivebehavior:a review. Psycho1Bull 1985;98:53&613. 3. Rogel MJ, RuehlkeME. Adolescent contraceptivebehavior: influencesand implications.In: S zuartI, Wells C, eds, Pregnanc,~in adolescence:Needs, pnDblems,and management. Nc..lrYork, Van Nostrand, 1982:194-216. 4. Clark SD, ZabinLS, Hardy JB. fix, contraceptionand parentheod: experienceand attitudes among urbanbiackyoung men. Fam Plann Perspect 1984;16:77-f32. 5. Cohen DD, RoseRD. Maleadolescentbirthcontrolbehavior: the importanceof developmentalfactorsand sex differences. J Youth Adolesc 1984;13:23%-52. 6. Dryfoos J. Contrxeptive use, pregnancy intentions, and pregnancy outcomes among U.S. women. Fam Plann Perspect 1982;14:81-94.
Nove
7. 21.
8. 22. CobIiIler WG, sch ceptive failures. J Psycho1 1976;13:23923. .@I&& B, Qskamp S. IndividuaI di&. W-K cent contraceptors. &n:Stuart I, We?: ‘I;, e adokxcence: needs, pobIems, and rnart!,~~
9. xl. 11.
24.
Ii.
25.
13.
26.
14.
27.
Satisfaction with health care: a D
28. 15.
29. Youth Adolesc 1980;9~49%-X%.
i6.
30. 31.
17.
32.
as.
19. Litt IF, Cuskey WF& Rudd S. Identifying adolescenti at risk fooxcocolIiance with contraceptive therapy. J IMiatr ,: . 20. Foreit JR” FoRit KG.
and contEacepf:ivel-U?hiW-
33. Hein K. Issues in adolescent heaealtian overview. New York, Carnegie Corporation, 1 s: an 34. Gates W, Rauh JL,. AdoIescents and sexu I-IeaIthCare a RatioEaIe for Concem. New 35. 988.