Tobacco Documents Online

I I I I I I I 1 I I I I I I I I I I I Addtetive Behavio.s. Vol. 15, pp. 235-245. 1990 0306-fi03190 S3.00 * .00 Pnaled w the USA. All ntfm neserved. Copyngtft 0 1990 Peraamoc Press plc SLIP-UPS AND RELAPSE IN ATTEMPTS TO QUIT SMOKIIVVG RON BORLAND Csatre for Behaviomal Raeazzh ia Csaca, vieeona. Ausnalia Aiaarftn - Ybis paper ts ooooetaed with docameat>nS tbe ooatextx ia .vh>c6 tiiQ~fpa ia P - to atop smoicm= occur. and of tlie mmequeaees of tbe slsp-up oo tbe omamaooo of amc+kmg oesaaaon. A saanpk of people who had called a Quit Stmiciaj w3epbooe aaviee for iafortaaaoe was txootacted ttuee mooths lamr. A eocal of mitul 216 alip,up eptsodes was tspoeed. vt of whub tssulted m a tesumpooo of abc>zaeacs and 172 wEictt led so Rlapae. Stip•apc wste found to ocaa ia a broad rantc of comxts, and coatext was also telated io outrome. Slip-nps when m a potiave tflood, when aocisliaa= and drmtsa= alcohol were associated witL tiislses kvels of taeovery of absanence. as was those thas ox>asred aHer moct daa two we¢!ts of abtaaesa. Sex di}fetmaes weze also fouad. with tnen moas likely m attp-ap at work. and women to sltp-ap ia a brwder vsnety of cootexts. A large amount of time and effort has been put into developing methods to help smokers give up their habit. In general, the results of these endeavours have not been impressive (Schwartz, 1987). In Australia, 85% of current smokers have made attempts to quit (Borland & Hill. 19884. It is clear that most quit attempts end in relapse. Such findings have led to a change in research emphasis. with some investigators beginning to focus on relapse episodes (e.g.. Marlatt & Gordon, 1980, 1985; Shiffman, 1982; Shiffman & Jarvik, 1987). The aim is to develop methods to reduce the occurrence of relapse. In the process of addressing this goal it is imporcant to obtain good descriptive information on the situations in which cigarettes are smoked after a period of abstinesxx (called slip-ups in this paper) or in which temptations occur which are overcome. Research into slip-up episodes is of two kinds, those that have provided a multivariate description of slip-up contexts (e.g., Cummings, Jaen, & Giovino, 1985; Shifhaan, 1982) and those that have used the set of mutually exclusive categories developed by Marlatt and Gordon (1980) (e.g., O'Connell & Martin, 1987). It is difficult to directly compare the two methods; they will be dealt with separately. Descriptions of slip-up episodes have been collected from both those who recovered from the slip-up but promptly resumed abstinence, and those for whom the slip-up represented the beginning of a>zlapse. In addition, several authors (e.g., O'Connell & Martin, 1987; Shiffman, 1982) have also studied crisis episodes which did not result in a glip-up (variously called close-calls or near lapses). Shiffman (1982) found that crises involving alcohol teaded to result in slip-ups, and that crises that did na end in slip-ups were associated with a broader range of cognitions. However, there is a major methodological problem in comparing close-calls with actual slip-ups when they are both recalled retrospectively. Crises that do not result in a slip-up can only be defined by the individual through the thoughts they have, whereas slip-ups can be identified afut the event independent of any thoughts had at the time. This means that it is impossible to report on a resisted temptation without having cognitions about it, while it is possible to have a slip-up with no associated This t=atsh .vas ftmded by the Aao.Caaeer Camcii of Vittoeia and the Victoeiaa Smokiat and Heahh PtoStam. T'hanks ars due to David }iill and VirSiaia Lswi: for ooatmeats on an eariier vasion of this majatsatpt. Raquesu for tepemts sbould be aett a Roe Bociutd. Ph.D.. Behaviomal Scieaast. Centre for Behaviourai Raeuctt m Caacer. Aam-Caooer Conncil of Viesona. 1 Radsdowne Street. Grltm South. Vicsona 3053, AnsQalia. 235

I I I I I I I I I I I I I I I I .36 RO': BORL.1.ti'D cognitions. This problem is likely to be maximal in Shiffinan's (1982) study. as his subjecu had telephoned a relapse crisis line. Even in studies wh.re the researcher selecu the subject, the problem is still likely to exist. There is likely to be an overszpresentation of situations with salient cues, either cognitions or subjective expetiencxs, among the close-calls as compared with slip-ups. because the slip-up itself is sufficiently salient to be reported independent of other cognitions. As a result, close-calls cannot be directly compared with actual slip-ups when using subject-defined criteria of crises. Slip-ups are likely to occur equally throughout the working day (Shiffatan. 1982). although Mariatt and Gordon (1980) found more slip-ups in the evening. Slip-ups have been found to occur most commonly at home, and in the presence of other smokers. Alcohol consumption has been found to be a factor in around one-quarter of cases, and consuming food and/or coffee is also common around relapse. Most relapses seem to occur when the person is in a negative mood, and to be ptecipitated by some kind of stressful crisis (Cummings et al., 1985; Shiffman, 1982). The generality of these results is open to question. Shiffman's (1982) data is from callers to a relapse hot line, and Cummings et a1. (1985) used people atroending a smoking cessation clinic. It is possible that the results typify smokers who find quitting especially difficult. Similar research is needed on samples who are moce characttristic of smokers in general. O'Connell and Martin (1987) have attempted to identify differences in slip-ups as a function of the outcome. They also included crises that did not result in slip-ups but these cases have been excluded from the results summarized here. They found few differences between temporary slip-ups ("lapses") and relapses. Relapsers were more likely than lapsers to report withdrawal symptoms, and interpetsonal conflict as the cause of the slip-up. Cuny. Marlatt, and Gordon (1987) studied 36 slip-ups, 20 of which resulted in a resumption of smoking (zzlapse). Tbey found that relapsers reported higher levels of what they call the Abstinence Violation Effect (AVE) than those who simply lapsed and resumed abstinence. The AVE is a central component of Marlsat and Gordon's (1980, 1985) theory of relapse. The AVE is the cognitive-emotional response to the initial slip-up and has two components: (a) a causal attribution of responsibility for the slip (personal vs. external to the individual), and (b) an affective reaction to the uaibution (e.g., guilt. feelings of failure). Mariatt and Gordon ptvpose that the intensity of the AVE is increased by causai attributions that focus on internal, stable and global factors that are perceived to be uncontrollable. Emotional reactions (e.g., guilt and self-blame) to such attributions inhibit recovery following a slip-up. This research suggests that if a method could be found to make AVEs more external, unstable, specific and controllable (i.e., due to circunnssaaas that the petson believes they can controi), it might be possible to reduce tbe probability of relapse following a slip-up. Prior to a more detailed investigation of the role of AVEs it is important to fusttter chatactesise the contexts in which slip-ups occnr becatue the context of the slip-up may affect the kinds of cogaitions that the person has. just as Martatt's tLeory would predict. The major aim of this paper is to provide descriptive information about the context in which a person smokes his/ber first cigarette (or puff of a cigareme) after a decision not to smoke again. The sample chosen is presumed to be closer in character to the general population of smokers who are plaaaing to quit than those reviewed above. Slip-ups following a quit attempt can either result in a full relapse. that is, a:eturn to smoking, or they may be transitory with the person renewing his or bet coaimitment to abstinence. A secondary aim is to consider whether the outcome of the slip-up (resuming abstfnentx vs. relapse) is associated with the context in which it occurs.

I I 1 I I I I I I I I I I I I I I I ReLqqe m qnimng unotint 237 METHOD SlcbjCctS The subjects were drawn from a pool of callers to the Quit Inforntation Line, a counselling and information service available during the major annual anti-smoking campaign in the state of Victoria, Australia (Victorian Smoking and Health Program, 1986, 1988). A sample of 536 smokers who intended to quit and who had agreed to a follow-up phone call was identifted from 711 callers in the last three weeks of a 12-week campaign, prematurely Ieaving a total of 378 respondents. A total of 378 of the 536 (70.5%) were interviewed about three months after the initial contact. The 158 unsuccessful calls included four refusals, four who were unable to complete the interview, 78 cases where the person no longer appeared to be at the address given, and 72 cases where the persott could not be contacted after at least four attempts. There were 149 males and 229 females of median age 35. These were predominantly white collar workers (54.596), with 19.8% blue collar workers and 25.7% not in the full-time paid workforce. In comparison with community estimates, the sample overrepresents women and people in the white collar sector of the workforce. The sample reported smoking an average 25.7 cigarettes a day at the time of initial contact. The structured interview Respondettts were taken through a structuted interview over the telephone. They were initially queried about whether they had stopped smoking for at least a day since calling the Quit Information Line. Those who had were queried about whether they had had a slip-up episode. All respondents who reported a slip-up were questioned about the context of the initial slip-up. In particular, they were asked: how long they had been abstinent at the time of the slip-up; what caused the slip-up; what they were doing; where and when it occurred; whether other smokets were present; where they got the cigarette: and what their mood was. They were also asked about how many cigarettes they smoked in the slip-up and whether it led to their resuming regular smoking. RESULTS The 378 respondents consisted of 64 persons who were no longer smoking. 25 of whom had had a slip-up episode, 191 respondents who stopped for at laast a day, but had resumed smoking by the time of the interview, and 123 who had not even stopped smoking for 24 h. Sixty-eight petrent of respondents tberefore made a quit attempt, including 16.9% who were abstinent at call back. Of the 255 sespondents who had stopped for at least a day, 216 reported slip-up episodes. Of tbese, 172 resulted in relapse (Relapsers) and 44 cases resumed abstinence (Lapsers) of which 25 were still abssinent at interview. It is apparent from these data that most slip-ups (80%) end in relapse. It is important to determine whether the contezt in which the slip-up occurs affects the rate of relapse. There was no relationship between the reported number of cigarettes smoked a day at initial contact and the outcome of the quit aaempt. Slip-up contexts and their relatronship to relapse The estimates of the petiod of abstinence prior to the slip-up were divided into three categocies: periods of 1-3 days, periods of four days to two weeks, and those of more than two weeks. Forty petoent of slip-ups occurred within the first three days of abstinence, another 34% within two weeks, leaving only 26% of ca,ses where the period of abstinence was longer than two weeks. There was a marked relationship between the length of the period

I I I I I e I I I I I I I I I I I I _38 RON BORL.A.`'D Table 1, Slip•up coatsxt by outcome hpsets Relapsers Sig Penod of absunence - y7-3 a3~s - 1496 4896 4 days-2 weeks 36% 34% X= = 23.1. df = 2 morc elsaa 2 weeks 50% 19% p<.0001 Reaon ~4ibElemvcnsu 26% 44% Mennl atutude 5% 15% x= = 21.3. df = 3 Addicaoa/babtt 19% 23% p < .001 Smotiat cues 51% 18% Where IW% 21% 33% Home 18% 42% z' = 23.4. df - 3 Social suuaaoas 46% 15% p<.0001 Otha 16% 9% Whea ~10fbnu0g 11% 35% Afranoon 46% 37% X= = 10.0 Eveaing 43% 27% p < .01 M d oo 'Miove 54% 26% Stssssed 12% 24% X= - 14.5. df - 3 Negaave 33% 36% p < .01 Other 2% 14% Who with IPM 67% 49% Noosmohers 14% 17% z= - 4.9, df = 2 Alooe 19% 34% NS Aeuvay ' ElMiak 9% 12% AkoW 41% 12% Relaxiag 5% 15% X2 - 21.6. df = S Woeiaag 23% 33% p < .001 Talkin= 11% 9% Otber 11% 19% Cigaese ftom - "ME! 71% 42% Had it 10% 36% z2 - 13.8 Boa& it 19% 22'R p<.001 of abstinence and the outcome (x2 = 23.1. df - 2. p<.0001). The eariicr the slip-up occurred, the mose likely it was to end in a full blown relapse (see Table 1). Reasons given for the initial slip-up were divided into four broad cuegoties. The tr.ason were crises, personal probkms and external life presslu+es (collectively calkd •'problems,'• accounting for 40% of all cases); smoking cues and situatioos (cues, 25%); a+ddiction, habit or transieat negative feelings (craving. 22%). and lack of willpower aad other aspects of mental attitude (13%). 'Ibore was a significaat overall relationship between reasons given aad the outcoate of the slip-up Q2 = 21.3, df = 3, p<.0001). As can be seen from Table 1. the difference was largely due to lapsers (as compared to relapsers) slippiag-up more in ~ ~ ~ ~ ~ ~

I I I I 1 I I I I I I 1 I I I I I RelaQae m qnza; smoi=e 239 the presence of smoking cues, and somewhat less in response to problems and mental attittide. . Slip-ups were reported as ooansing at home (37%), work (30%) and among social contacts, largely in nds' homes (21%), with only 11% in other situations (e.g., in urs, shopping centres. outdoors. etc.). Again there was a difference as a function of otucome. ()(2 = 23.4. df = 3. p<.0001), with Lapsers being motz likely to slip-up in social situations and Relapsers to slip-up at home (see Table 1). Slip-ups were reported as being fairly evenly spread across the day with 30.0% occurring in the morning, 39% in the afternoon and 3196 in the evening. Again there was a relationship with outcome (x2 = 12.2. df = 4, p<.05). Lapsers were more likely to slip•up luer in the day than were Relapsers. The petson's mood was also related to slip-up; 32% were while in positive moods. 35% in negative moods (sad. angry, upset. anxious), 21% when under stzess or ptr.ssures, and 12% in an "other" category that largely consisted of less common negative mood states. There was a significant relationship between mood and outcome (X2 = 14.5, df = 3, p< .01). Lapsers were tnore likely to slip-up when in Positive Mood states. The majority of slip-ups occurred in the presence of other smokers (53%), compared to 16% in the company of nonsmokers and 31% while alone. Outcome was unrelated to the company kept during the slip-up episode. The most common activities reported to be associated with slip-ups were work (3196), drinking alcohol (18%), relaxing (13%), eating and/or nonalcoholic drinking (12%) and talking (9%), leaving a large 18% in a variety of other situations such as driving a car and waiting. Activities were significantly related to outcome ()(2 = 21.6, df = 5, p < .001). Lapsers were more likely to slip-up while drinking alcohol. Slip-ups which result in relapse appear to occur with a broader range of activities (see Table 1). The most common source of cigarettes for a slip-up was other smokers (either offered or requested, 48%), while 31% reportad having cigarettes or around about, leaving only 21% who had to buy them. Outcome differed as a function of source of cigarettes. (X2 = 13.8, df = 2, p<.001). Lapsers were more likely to get their cigaretus from othets. Taken together, there is an intpressive amount of difference between Relapsers and Lapsers given the limited power of the study. Covariarion of aspects of the slip-up contrzs Thete were sttoag interactions between the various context variables. To give a picture of this interaction, associations with each of the categories of action engaged in and kngth of period of abstinence have been explored. Tbet+e were highly significant associations between the activity at slip-up and each of the otlter seven context variables. The initial period of abstinence was associated with the activity at slip-up (X2 = 23.9. df = 10, p<.01). Alcobol-related slip-ups. tieaded to occur after a longer period of abstinence while the others either showed no pattaa, or tended to decline with the length of period of abstinence (see Table 2). The main trason given for the slip-up and the activity engaged in were highly associated (X2 - 93.8, df - 15, p < .0001). Forty-four petrent (n = 37) of slip-ups attributed to problems were at work, as were 36% (n = 10) of those atuibuted to mental attitude. By coacast, 55% (n - 29) of slip-ups attributed to smoking cues occurred when drinking alcohol (see Table 2). The place of the slip-up varied with the activity, (X2 = 285.3. df = 15, p<.0001). The effect was mainly due to 60% (n - 15) of eating and drinking related slip-ups being at hotne, as were 9196 (n E 32) of tltose when relaxing. Seventy-eight percettt (n = 31) of slip-ups involving alcohol were in social situations, and not surptisingly, 8296 (n - 62) of

I I r I I I I I I I I I I I ~ I 2.t0 RON BORLAND Table 2. Numbers of slip-ups a funcuon of "Relapse Context" Situation Eav drtnk Akohol Relaxin8 Workin8 Taliin8 Otl+er Persod of absatxnce - y 14 6 13 30 9 12 4 days-2 weeks 6 14 10 22 4 14 ssorc ttun 2 weeks 5 18 2 13 7 7 Reasott "'Pfobiems 8 4 8 37 11 16 Mentai atntude 1 3 5 !0 0 9 Addictiott/habit 11 2 10 13 2 10 Stnot,in8 0ues 5 29 4 6 7 2 Whese "'R'&t 4 1 1 62 4 3 Home 15 6 32 12 9 19 Soeul 5 31 1 0 9 2 Ckhet 1 2 1 2 1 18 wAen 10[SrnmB 11 2 11 30 6 14 Aftertwon 8 9 11 38 6 16 Eventn8 7 29 12 5 8 5 Mood mit9ve 10 32 11 10 5 5 Stsess/Posiuve 2 2 5 27 3 10 Nefanve 9 5 14 29 12 17 Otber 4 1 4 9 2 7 Who with '%W;Brcis 12 37 8 35 17 6 Non•Imroitets 5 1 8 16 5 5 Alone 9 2 18 21 0 23 Ciaatette from "'WES- 9 30 6 37 17 4 found tt 10 6 24 23 3 21 Bouiht it 6 2 5 15 1 14 work-related slip-ups were at work. No place predominated for talking or other activity (see Table 2). Tbe tiax of slip-up wu telated to ttte acciviry (X2 = 62.6. df z 10, p<.0001). T}x main sssocistions wRere alcohol consumption in the eveaings, (7396, n- 29) and both work and other, which rarely happened in the evening (7%, n= 5, and 1496, n- 5 respectively). Mood at slip-up was also related to activiry (X2 - 74.3, df = 15, p<.0001). Mood u slip.up was likely to be positive when driaking alcohol (80%, n= 32). and slightly mote likely to be negative when talking (5596, e= 12), while in otber contexts tbem was no clar pan=''bee social context of the slip-up was also associated with activiry (X Z = 71.3, df - 10. p<.0001). Slip-ups tended to occur with other smokers when drinking (93%. rt - 37) and t;tllcitlg (7796. R= 17), while telaxing (5396. n - 18) and otber (6846. n= 23) teaded to oxur when alone. As might be expected from this pauern, tisets was also a relationstup between activity and ttle3ounce of the cigusae (X= = 67.2. df - 10. p<.0001). Drinkers (7996, n= 30) and talkers (81%. n= 17) got their cigarettes from others. while those C~3 ~ ~ CO O 1~

I I I RdWpse in quitang smotsng 241 Tabk 3. Length of iaitial abadneoa by oooKxt of slip-up I I I I I I I 1 I I I I I Length Corttsxt 3 days days- ocs tun 2 wroeks 2 reeks Significance Resson "rWems 37% 35% 5M blensai amwde 16% 13% 6% x2 - 23.2 Caving 33% 23% 6% df - 6 Cu" 14% 30% 39% p<.001 Where Cork 41% 24% 25% Home 44q6 42% 17% X2 - 30.5 Social 99E 23% 42% df - 6 her 7% 11% 15% p<.0001 When WSening 3896 27% 22% X, - 12.5 AftRSfoOCf 46% 35% 37% df = 4 Eveaing 17% 38% 41% p<.05 who with ' Sf4Wcrs 399E 55% 75% x' - 16.5 Non-smoicas 22% 13% 10% df - 4 Others 40% 31% 16% p<.01 Ciguena from ' - OR is 32% 599E 65% x2 ~ 22.8 Bought it 43% 25% 10% df 4 Had it 25% 16% 25% p<.0001 relaxing tended to have them lying about (69%, n= 24). The length of the period of abstinence was sig,nificantly associated with all other context variables except mood (see Table 3). The major effects were that craving became a less common reason and cues a more common reason over time. Slip-ups were more likely to occur at work in the first three days and at home in the first two weeks, while they were increasingly likely to occur in social situations as the period of abstinence increased. Slip-ups were also less likely to occur in the evening during the first three days of abstinence. They tended to occur more when alone in the first two weeks and increasingly in the presence of other smokers over time. In parallel with this, the slip-up cigarettes were increasingly obtained from others, and less likely to be had on or about the person. One other tlntabulsted associuion warrants mention. Mood was associated with time of day (X2 = 45.0, df = 6, p < .0001). Most positive mood slip-ups occurred in the evening (51%), most stt+ess-telated slip-ups in the afternoon (63%), with very few in the evening (491c), and most "other" mood slip-ups in the moming'(6396). Negative mood slip-ups were reasonably evenly distributed across the day. One constellation of factors appears to stand out. It is that alcohol-related slip-ups occur mostly when the person is in a good mood in social situations, in the evening, with other smokers present from whom the person gets his or her cigarettes. However, the prognosis from these slip-ups is the most optimistic as this combination of context variables is also associated with high levels of recovery from the slip-ups. Sex differences arwng slip-ups There was no significant sex difference in the outcome of slip-ups. nor in the proportions of males and females having slip-ups. Further, ther+e were no sex differences in length of I

I I I I I I I I I I I I 1 I I I r 242 RON BORLAND Table 4. Relapse context by sex (si;nifwnt effecu) Male fesaak Sitaifxcanoe wl+ett -oric Ia% 19% x= - 23.4 Homa 24% 46% df - 3 Social 22% 21 % p < .0001 Otber 6% 14% Acaviry "'EW'lirink 8% 14% Drink dcolwi 25% 14% X2 - 23.0 Relwa= 11% 14% df - S Wod 44% 23% p < .p01 TAlhint 5% 12% Otber 8% 24% Mood To-sitive 38% 28% x= ~ 9.5 PressurrlSoeu 27% 17% df - 3 Nepcve (sQecifiod) 24% 43% p < .05 Other 11% 12% Cieuette fraa TOMs 61% 40% X= ~ 9.6 Had/found it 20% 38% df ~ 2 Souftu it 20% 22% p<.01 initial period of abstinence or in reasons for slip-up, but sex differences were associzued with several other variables (see Table 4). Although the time of day the slip-up occurred was unrelated to sex, there was a sex difference in where the stip-ups occurred (XZ = 23.4, df = 3, p < .0001). Men were more likely to report slipping-up at work (48%), while the women slipped up more at home (46%) and in "other" situations (14%). The activity at slip-up also differed (X2 = 23.0, df = S. p<.001). with men more likely to slip-up when working and when drinking alcohol. and women care likely to slip-up in "other" situations. In general. it appears tllat women slipped-up in a broader variety of situations and places than men. It should be noted that the difference in slip-up as a function of alcohol drinking disappeued when those not in the paid workforce were excluded from the analysis. Mood at slip-up was a1w eelared to the sex of the smoiter (X= a 9.5, df = 3. p<.05). Women were moce likely to raport negative moods ar slip-up, particularly sadness and depression. The other significant effect was in the source of the slip-up cigarette Q2 - 9.6, df = 2. p<.01). Women were mote likely to have had the cigarette on them or found it ne:rby, and the men tnore likely to get it from somebody else. There was no cleu evidence that any of those differences had any impact on the outcome of the slip-up. DISCUSSION 'tbe results reported in this study are generally similar to those in other compmble studies (Cummings et al., 1985: Shifftaaa, 1982) but exoead them in tbree major respects. Firstly. the results generally confirmed the repocted patterns of slip-up occurrences in a sample which is closer to the geaaal population of smokers intending to quit and extended them by identifying work as an important locus of slip-ups. Secondly, the study identified clear sex diffeteaces in the contexts in which slip-ups occur. Thirdly, the results provide evidence that the context of the slip-up was associated with the likelihood of resuming the quit attempt. ~ ~ W Ca ~ Q ~ ~

I I I I a I I I I I I I I I Relap.e m qwmag smoi=t 243 T6e pattern of slip-up coatexts was similar to that reported in studies by Shiffman (1982) and Cummings et a1. (1985). The only notable difference was that in the present study, work appeared to be a more important perceived cause and place of slip-ups. It may be because people in the workfo:ze were underrepmsented in the previous studies. This could be because working people have less time to become involved in intensive smoking cessation programs and/or they have less nead for the social support such progranu provide. Whatever the teason for its absence as a major factor in past studies, this study clearly shows work to be a major Iocus of slip-ups, especially among men. Tbe sex differences fottnd in this study are of potential relevance to campaigns to facilitate quitting. Women appear to slip-up in a broader variety of situations, and seem more likely to be experiencing negative emotions when they do slip-up. Men, by contrast, mote often report slipping-up when under external pressure, and consistent with this, they slip-up mots often at work. It may be that smoking is playing slightly different roles in the lives of male and female smokers, with men more likely to smoke while dealing with external pressures, and women to smoke in response to the emotional consequences of such pressures. It may be that this difference is associated with differences in the optimal coping strategies. Slip-up crises thu occur when engaged in activities (e.g., work) may be more difficult to deal with at the time because of the competing demands of the ongoing activity. Unless the maintenance of abstinence is given a higher priority than the immediate activity. the possible options for successful coping will be limited. In such cases a slip-up may be unavoidable. If this were so, it could be important for these episodes to have prepared strategies during the early part of the quit attempt, e.g., like making provision for work demands to be moderated. In cases where there is no competing activity, the person may have more opportunity to direct his or her energies towards overcoming the temptation to smoke. Further, they may have a greater range of alternatives open to them. For example, take the tisk associated with drinking with fr[ends. There are many options which include avoiding the situation, asking the friends to provide support, and reducing alcohol intake. This means that there is a good chance that slip-ups can be avoided. By contrast, when there is a work deadline to meet it is difficult to maintain uaying stopped as the number one priority, thus limiting the options for coping, and making a slip-up more likely. Of particular note is the finding that people attempting to quit were more likely to recover from slip-ups associated with socializing and drinking. This contrasu with Shiffman's (1982) finding that such situations are more likely to result in slip-ups than abstinence, but it is not necessarily cottttadictory. It may be that the actual slip-up is mom likely when socializing, but, perhaps because of the perceived external presstnrs, it is easier to recover from such slip-ups. It is apparent from the results that the longer the period of abstinence, the mote likely it is that the individual will recover from the slip-up. Berause a resumption of abstinence becomes more likely as the period of initial abstinence incmases, it is possible that some of the differences in contexts may be a function of the period of abstinence as ex-staokers test themselves in progressively more difficult contexts. For example, alcobol consumption is a slip-up context that appears most important the first two weeks after cessation. One explanation of this would be that people adopt the sensi(zk strategy of avoiding drinking situations early in the quit attempt. Alternatively it might be that vigilance about risky situations decreases with the length of the period of abstinence. Shiffman (1982) found some evidence for this latter explanation. Regardless of the mechanism by which some situations are managed early in the quit attempt. there is a need to focus on the skills required to cope with the situation in which early slip-ups commonly occur. This is because no attempt can succeed unless those immediate high-risk situations are survived. Furthermote, slip-ups early in the attempt are

I I I I I i I I I ~ I I I I I I I 244 RON BORL.&%'D likely to have a more marked negative effect on self-efficacy beliefs. This may inhibit persisting with the attempt, and when this occurs, further reduce the likelihood of trying again in the future. The relationships between the context of the slip-up and its outcome are generally consistent with Marlatt and Gordon's (1980. 1985) theory of AVE as a determinant of relapse. Where aspects of the situation are more likely to result in utribution of external causation (e.g.. drinking, getting the cigarette from others), and thus should result in less intense AVEs, recovery from slip-ups was more common. Similarly, where affect was positive, and thus the AVE was less likely to be sevete, recovery was more frequent. The major discrepency between predictions based on the AVE ttuocy and these results was the lack of effect on outcome as a function of the presence of other smokers. The subjective abstinence violation effect might be expected to be less intense in cases where other persons were available to blame for the slip-up, that is. external amibutions of causality would be more probable. The finding that longer periods of abstinence was associated with higher recovery is compatible with the role of self-efficacy (Bandura, 1977, 1986) in moderating the AVE. The evidence the person gains from the period of abstinence that he/she can remain abstinent for extended periods seems to increase her/his tendency to have another go (resume abstinence) after a slip-up. Taken together. the results clearly indicate that the context of a slip-up is telated to its outcome. It is important to consider whether some situations are intrinsically more difficult to recover from than others, or whether the differences are due to other factors. It was argued above that in situations where there is a high priority for ongoing accivity, psychological resources will be less likely to be available to add:ess the crisis so relapse may be nsose likely. However, if smokers are to succeed in quitting their habit, they must adopt strategies to cope with such situations. Forward planning and leaming from mistakes would seem to be important strategie! to use to ensure that the cognitions and behaviours needed to maintain abstinence can be implemented when necessary. This means that any situations which lead to potentially competing behaviours and cognitions must be regulated. This may involve negotiating with others to assume responsibilities or to assist in pressing tasks, so that these pressures do not take priority over the cognitions and behaviours required to tnaintain abstinence. It was suggested above that avoidance of some high-risk situations is a stratsgy that some quitters adopt early in smoking cessation attempts. This is a sensible strategy as it can free up resources to cope in the high-risk situations that cannot be avoided as readily. Further, it appears tbat. at least in the case of drinking alcohol, avoidable contexts which lead to slip-ups are more likely to be recovered from. Interpreted in teerims of Marlaa and G cdon's (1985) AVE. this may be because avoidable coatextt are mo[e likely to be perceived as external, uasuble and controllable by the iadividual. Thus there is likely to be kas need of professional assistance to help the pesoa understand that these situations are ones that be or she can influence. By contrau, unavoidable situations often result in relapse. Unless people can Set through the s=ona that ttfey cannot avoid, ttiey are never going to need to face the ones they can. 'Ibe focus of research amatioa needs to be directed towards a betser tmdessraadiag of the common unavoidable slip-up coatextu, so that better stiaugies can be suggested as means of overcoming the tempation to smoke in tbese situarions, or to facilitate recovery of absrinenee when they occur. One of the major, relatively unavoidable situations in which relapse is common is in copingwit3iitressful eveats. As.vss noted above, some saessors (e.g., work) can be regulated by gaiaing assisonce and support fram others. Further. anxiety and teasion can be managed through the use of relaxation sechniques or by snvcnuing time-outs into the Z1Z ~ ~ ~ C.~ CA O ~ C3t