Coping with psychotic symptoms in the early phases of schizophrenia

American Journal of Orthopsychiatry, 70(2), April 2000

© 2000 American Orthopsychiatric Association, Inc.

Coping with Psychotic Symptoms in the EarlyPhases of Schizophrenia

Steve Boschi, M.S.W., Richard E. Adams, Ph.D., Evelyn J. Bromet, Ph.D., Janet E. Lavelle, M.S.,Elyse Everett, M.S.W., Nora Galambos, Ph.D.

Hoo people diagnosed with schizophrenia cope with positive symptoms after theirfirst

hospitalization is explored, along with the relationship of their coping strategies to their

psychosocial functioning. The strategies mostfrequently endorsed were cognitive in type,

while those considered most helpful were behavioral. Respondents identifying an active

strategy as most helpful displayed better psychosocial functioning at 24-month followu-up.

S ince the start of deinstitutionalization in themid-1960s, the maintenance of individualswith schizophrenia in the community rather

than in a hospital has been a primary therapeuticgoal. While antipsychotic medications may reducethe severity of the psychosis, many such individu-als continue to experience delusions or hallucina-tions (Nayani & David 1996) and must learn tocope with psychotic symptoms if they are to attaina stable life in the community. Several studies haveshown that individuals suffering from mental ill-ness try to control their symptoms through variouscoping strategies, including active-behavioral, ac-tive-cognitive, and avoidant techniques (Breier &Strauss, 1983; Carter, MacKinnon, & Copolov,1996; Cohen & Berk, 1985; Falloon & Talbot,1981; Lee, Lieh-Mak, Yu, & Spinks, 1993; Romme& Escher, 1989; Romme, Honig, Noorthoorn, &Escher, 1992). Nayani and David (1996) found aninverse relationship between the number of copingstrategies patients employed and the degree of dis-tress they experienced from their symptoms. Inevaluating different styles of coping with auditorysymptoms (e.g., hearing voices), Farhall and Gehrke

(1997) found that acceding to the voices (active ac-ceptance) was associated with a sense of control,reliance on external sources (passive coping) wasassociated with less subjective distress, and prob-lem-focused or active-behavioral (resistance cop-ing) was associated with increased level of distress.

Other than gender, the demographic correlates ofcoping have rarely been studied. The findings onsex differences in number or type of coping strate-gies have been inconsistent (Cohen & Berk, 1985;Lee et al., 1993; Carter et al., 1996; Falloon &Talbot, 1981). For example, two studies (Carter etal., 1996; Falloon & Talbot, 1981) found no mean-ingful differences by gender. In contrast, Cohenand Berk (1985) reported that women were morelikely to use prayer or seek medical services in re-sponse to symptoms than were men. Lee and col-leagues (1993) reported that social contact experi-ences were more helpful for women than for men.

Another little studied issue is whether certaincoping strategies are more effective than others inimproving patients' psychosocial functioning. On-ly two studies, both utilizing outpatient samples,appear to have addressed this question. The first

A revised version of a paper submitted to thle Journal ill June, 1999. Work was supported by grant MH-44801 from the NationalInstitute of Mental Health. Authors are at: State University of New York at Stony Brook (Boschi, Bromet, Lavelle, Everett, Galam-bos) and Mount Sinai Medical Center (Adams).

242

Boschi et al

(Falloon & Talbot 1981) found no systematic dif-ferences in types of coping strategies among pa-tients who had adapted well to their hallucinationsand those who had not The second study (Lee etal. 1993) also found no significant relationship be-tween types of coping and psychosocial function-ing, but did find that the total number of strategiesemployed was positively associated with social ad-justment, quality of life, daily life adjustment, andsymptomatology. Effective coping, therefore,should be related to improvement in current symp-toms and functioning, as well as to future symp-tomatology, psychosocial functioning, and qualityof life.

The study reported in this article explores 1) theways in which patients with schizophrenia copedwith psychotic symptoms during the six monthsfollowing their first hospitalization; 2) the degreeof distress caused by hallucinations and delusions,and the degree of control over them perceived byrespondents; 3) the coping methods perceived tobe most helpful by respondents; 4) the relationshipof gender, race, symptomatology, and quality oflife to coping strategies at the end of the sixmonths; and 5) the predictive utility of coping withpsychosis at six-month follow-up with psychoso-cial functioning 24 months after the initial hospi-talization.

METHODSample and Procedure

The sample was part of a Suffolk County, NewYork, cohort of first-admission patients admittedprimarily to one of 12 inpatient psychiatric facili-ties in the county between September 1989 andDecember 1995 (Bromet et al., 1992, 1996).

Inclusion criteria for the study were: aged 15-60years; residence in Suffolk County; and clinicalevidence of psychosis, prescription of neurolepticmedication, and/or a facility diagnosis indicatingpsychosis. Exclusion criteria were: a first psychi-atric hospitalization more than six months beforethe current admission, moderate or severe mentalretardation, or an inability to speak English. A his-tory of drug or alcohol abuse was not an exclusioncriterion.

Most initial interviews took place in the hospital,shortly before discharge and after written informedconsent was obtained. Face-to-face follow-up in-terviews were held six and 24 months later, withtelephone contact every three months betweenbaseline and 24 months. The initial response rate

was 72%. Of the 695 patients interviewed at base-line, follow-up information was obtained for 650at six months (93.5%, of whom 88% were con-tacted directly) and for 581 at 24 months (83.6%of baseline, of whom 88.8% were contacted di-rectly).

Analysis focused on respondents who 1) re-ceived a longitudinal research consensus diagnosis(see below) based on DSM-IV (American Psychi-atric Association [APA], 1994) criteria for schizo-phrenia, schizoaffective disorder, or schizophreni-form disorder following a 24-month follow-up; 2)experienced delusions or hallucinations at anytime between baseline and six-month follow-up;and 3) completed the coping section of the six-month follow-up interview.

Of 234 respondents diagnosed with one of thethree types of schizophrenic disorder, 194 were in-terviewed face-to-face at six months, thus provid-ing reliable symptom information. Delusions orhallucinations during the prior six months were re-ported by 128 of the 194; 33 of these 128 respon-dents did not complete the coping module, primar-ily because interviewers were instructed to omitthis section if the respondent either denied psy-chotic symptoms or did not find them problematic.Thus, data on coping with psychosis were obtainedfrom a total of 95 respondents diagnosed withschizophrenia (76), schizoaffective disorder (17),and schizophreniform disorder (2).

Overall, the 95 respondents with available cop-ing data were predominantly 18-22 years old at

Table 1

DESCRIPTIVE CHARACTERISTICS OFRESPONDENTS WITH AND WITHOUT COPING

INFORMATION

COPING INFORMATION% WITH % WITHOUT

CHARACTERISTIC (N=95) (N=33)Age at Baseline (yrs.)

18-22 51.6 21.223-35 32.6 60.636+ 15.8 18.2

Race, black 24.2 15.2Gender, male 66.3 72.7Education, high-school graduate 66.3 75.8Ever married (at baseline) 21.1 15.2Living with family (at baseline) 82.1 60.6Using drugs/alcohol (at baseline) 51.6 45.5Onset to hospitalization <365 days 5 8 .7a 5 5 .2 bHospitalization at baseline <30 days 54 .9b 36.7 a

Antipsychotic medication (at 6 mos.) 80 .4a 78.8Consistent use of medications 54.8c 33.3 d

aThree missing cases.bFour missing cases.cEleven missing cases.dNine missing cases.

243

Coping With Psychotic Symptoms in Schizophrenia

baseline, white, male, high-school graduates, nev-er married, and living with their families at thetime of their first hospitalization (see TABLE 1).

MeasuresDiagnosis

The Structured Clinical Interview for the DSM-III-R (SCID) (APA, 1987; Spitzer, Williams, Gib-bon, & First, 1992), supplemented by questionsfrom the National Household Survey on DrugAbuse (1988), was administered at the three majordata collection interviews (baseline, six, and 24months) by master's-degree-level mental healthprofessionals. The follow-up SCID was modifiedto cover the intervals since baseline. Followingcompletion of the 24-month assessment, two pro-ject psychiatrists independently reviewed the threeSCIDs, all hospital discharge summaries, andother relevant information from respondents andsignificant others. Their diagnostic formulationwas presented at a meeting of all project psychia-trists, and a research consensus diagnosis wasreached (Fennig, Craig, Tanenberg-Karant, & Bro-met, 1994).

CopingAs an adjunct to the SCID module for psychotic

symptoms, respondents were asked a series ofquestions about how they coped with these symp-toms. Due to time constraints, the coping questionsreferred to any symptom, rather than specific onesexperienced by the respondent. There were 18coping items, modeled after the work of Moos andassociates (Holanan & Moos, 1987; Moos, Finney,& Cronkite, 1990), and respondents were askedwhether (yes/no) they used each strategy to dealwith their symptoms (see TABLE 2 for list of strate-gies). In addition, respondents rated how stressfulthe psychotic symptoms were for them (categor-ized for analysis as "not at all" or "somewhat" ver-sus "a lot" or "a great deal"), and how much con-trol they felt over the symptoms (categorized foranalysis as "none" or "somewhat" versus "a lot" or"a great deal"). Finally, interviewers inquired whichstrategy, if any, was the most helpful in dealingwith symptoms.

Again following the work of Moos and col-leagues, the 18 coping items were grouped intothree general coping styles: active-behavioral, ac-tive-cognitive, and avoidant (see TABLE 2). Previ-ous research (Holahan & Moos, 1987; Pearlin &Schooler, 1978) has emphasized that stressful

events elicit both active and passive (i.e., avoidant)coping strategies. Active coping strategies are at-tempts to alter the stressful situation itself or tomanage the resultant emotions. Avoidant copingstrategies, in contrast, do not deal directly with thesituation or the emotions, but reduce tensionthrough diversion (e.g., smoking or eating). Moosand colleagues further divided active copingstrategies into cognitive (e.g., try to remain posi-tive) and behavioral (e.g., talk to others) types.

All subscale scores range from 0 to 6 and aretreated as interval-level measures in the analyses.Finally, the strategy identified as most helpful bythe respondent was categorized as active-cog-nitive, active-behavioral, or avoidant.

Preadmission CharacteristicsData were collected on whether (yes/no) respon-

dents met DSM-III-R (APA, 1987) criteria for sub-stance use disorder at the time of first admission,the interval length (in days) from first psychoticsymptom to first hospitalization, and length (indays) of first hospitalization.

Medication at Six MonthsInterviewers obtained information from the re-

spondent, significant others, and clinical recordsregarding usage of antipsychotic medication overthe six months prior to the first follow-up. Fromthis information, two variables were created: wheth-er (yes/no) the respondent was taking an antipsy-chotic medication at six-month follow-up, andhow consistent ("not" or "somewhat" versus "very")respondents claimed to be in taking the medica-tion. Consistency in taking medication was basedon the clinical judgement of interviewers using allavailable information for the period between base-line and six months. This strategy appears to yielda reasonable approximation of medication usewhen compared to other methods. In previousanalyses of a subsample of this cohort, Gibson,Lavelle, and Bromet (1999) found that 94% ofself-reported medication was congruent with out-patient medical records. In addition, for respon-dents reporting dosage, there was an 87% agree-ment with outpatient summaries.

Psychosocial Functioning at Both Follow-UpsEight measures were employed to determine

whether coping strategies had a positive impact onpsychological symptoms, social functioning, qual-ity of life, and depression, as follows:

244

Boschi et al

1. The Brief Psychiatric Rating Scale-Anchoredversion, total score (BPRS-A) (Woerner, Mannuz-za, & Kane, 1988), which lists 18 symptoms, ratedfrom 0 (absent) to 7 (most severe) with a possiblerange of 0 to 126 (Lx=.74).

2. The Scale for the Assessment of NegativeSymptoms (SANS) (Andreasen, Arndt, Alliger,Miller, & Flaum, 1995). The SANS score repre-sents the average of five global interviewer ratings(affective flattening, alogia, apathy, anhedonia, at-tention) rated from O=none to 5=severe (cx=.80).

3. The Scale for the Assessment of PositiveSymptoms (SAPS) (Andreasen et al., 1995). TheSAPS score is the average of four global inter-viewer ratings (hallucinations, delusions, bizarrebehavior, thought disorder), using the same 0-5scale (ac=.67).

4. The Gl9bal Assessment of Functioning (GAF)(APA, 1987; Spitzer, Gibbon, Williams, & Endi-cott, 1996). The GAF was rated for the worst weekof the month before the interview. The possiblescore range is 0-90, with higher scores reflectingbetter social functioning and less psychopathology.

5. The Bradburn Happiness Scale (BHS) (Brad-burn, 1969). This scale is the sum of five items(yes/no), assessing the respondents' subjectivefeelings about how things are going in their lives,with higher scores reflecting greater happiness.The scale range is 0-5 (o(=.75).

6. The social functioning ratings of the Qualityof Life Scale (QLS) (Heinrichs, Hanlon, & Car-penter, 1984), an interviewer's assessment, basedon a semi-structured interview, of the respondent'slevel of functioning in certain social roles. The so-cial functioning subscale is the mean of nine items(oi=.93) rating participation in eight social activi-ties (e.g. relationships with friends, recreationalactivities, sociosexual relations) and overall levelof satisfaction.

7. The role functioning ratings of the QLS. Thissubscale is the mean of three items (Lo=.84) evalu-ating paid and unpaid work and level of accom-plishment. The range for both scales is 0-6, withhigher scores reflecting better functioning.

8. The Hamilton Depression Scale (HDS) (Ham-ilton, 1960; McDowell, & Newell, 1996) consistsof 21 items rated by interviewers to indicate sever-ity of depressive symptoms (o=.75). The scorerange is 0-65, with higher scores indicating great-er severity.

Intraclass correlations assessing interrater relia-bility for the symptom measures at six and 24

months, respectively, were .89 and .86 for theBPRS, .80 and .79 for the SANS, .82 and .92 forthe SAPS, .86 and .87 for the GAF, and .86 and .92for the HDS.

AnalysisAnalyses were conducted in three parts, using

bivariate and multivariate procedures. After de-scribing the characteristics of the sample and pre-senting descriptive statistics on the outcome mea-sures, contingency table analysis was used to as-sess relationships between individual coping strat-egies on the one hand, and gender, race, stress, andcontrol.

Next, Pearson correlations and t-tests were usedto analyze bivariate relationships among the threecoping styles, total number of coping strategies,active coping as most helpful, social characteris-tics, premorbid adjustment, and the psychosocialmeasures.

The final set of analyses focused on the relation-ship between identifying an active-cognitive oractive-behavioral coping strategy as the most help-ful in dealing with positive symptoms and psy-chosocial functioning at the 24-month assessment.More specifically, eight regression equations werecalculated with the 24-month psychosocial func-tioning measures as dependent variables and thesix-month psychosocial functioning measures andactive coping as most helpful as the independentvariables. These equations were designed to assessthe contribution of active coping strategies as mosthelpful to 24-month social functioning, controllingfor six-month social functioning.

It should be noted that some of the variables de-scribed above contained missing information. Thesix- and 24-month interviews were conducted intwo parts, with the QLS subscales and BHS in part2. Some respondents completed part 1, but failedto complete part 2. (See TABLES 1 and 3 for thenumber of valid cases for each variable, and TABLE

5 for the number of cases used to calculate the re-gression equations.)

RESULTSPsychosocial Characteristics

TABLES 1 and 3 present descriptive statistics forthe sample. At baseline, approximately half the re-spondents (51.6%) met criteria for lifetime sub-stance abuse, 58.7% had their first psychotic symp-tom less than a year before hospitalization, and54.9% were hospitalized for less than 30 days at

245

246 Coping With Psychotic Symptoms in Schizophrenia

Table 2

PERCENTAGE DISTRIBUTION OF COPING STRATEGIES

STRATEGYActive Cognitive

Prayed for guidance/strengthPrepared for the worstTried to see the positive sideDrew on past experiencesTook things one day at a timeTried to be more objective

Active BehavioralTalked with family memberTalked with friendTalked with professional personGot busy to keep mind off itTried to relax.Exercised more than usual

AvoidanceTook it out on other peopleAte/smoked more than usualUsed drugs or alcoholSlept a lotTried to ignore itKept feelings private

aRanking of strategy, based on percentage of respondents using it.

baseline. At the six-month assessment, more than75% scored below 41 on the GAF, more than 80%were on antipsychotic medication, and 55% werejudged by the interviewer to be very consistent intaking their medication. Finally, little change infunctioning was evident between the six- and 24-month follow-ups (see TABLE 3, column 1), exceptfor a decrease in depressive symptoms as mea-sured by the HDS (t=3.48, p<.001, dependent t-test). These findings indicated that, as a group, therespondents analyzed in this report were quite illand not functioning well at either assessmentpoint.

A series of comparisons on the same variableswere made to determine whether the 95 respon-dents analyzed differed significantly from thosewith recent symptoms who did not complete thecoping module (see TABLES 1 and 3). Applying aBonferonni correction of p<.002 (.05 divided by28 comparisons), none of the differences betweenthe groups was statistically significant.

Stress, Control, and Coping StrategiesA majority of the 95 respondents indicated that

their symptoms were highly stressful (56.8%, N=54) and that they had little control (70.5%, N=67)over them.

Of the 18 possible coping strategies, the meannumber endorsed was 9.75 +2.61 of a maximum of18. All respondents endorsed at least two strate-

gies, and 14 respondents used 13-15 of them, butno respondent endorsed all 18. Subjects endorsinga specific strategy (see TABLE 2, column 1) rangedfrom 12.6% who used drugs or alcohol to 85.3%who tried to relax.

Of the three coping style categories (shown inTABLE 2), respondents reported, on average, 3.65_1.40 active-cognitive strategies, 3.39± 1.32 active-behavioral strategies, and 2.71 +1.20 avoidantstrategies (see TABLE 3, column 1).

Most Helpful Coping StrategyMost respondents (86.3%, 82/95) identified a

particular coping strategy as most helpful (fivesaid that nothing helped, and eight did not specifyany single strategy) The most frequently cited in-volved talking with a professional (N=l l, 13.4%),getting busy (N=10, 12.2%), praying (N=8, 9.8%),and trying to ignore the symptom (N=8, 9.8%).Another 17.1% (14/82) named a coping strategyother than the 18 on the list (e.g. moving to a newlocation, quitting a job, taking medications).

Of the three coping style categories, 43.9%(36/82) of respondents claimed that active-behav-ioral strategies were the most helpful, 20.7% (N=17) cited active-cognitive strategies, and 18.3%(N=15) cited avoidant strategies. Overall, respon-dents nominated a greater number of active-cog-nitive strategies (e.g., trying to see the positiveside) when attempting to deal with their symp-

TOTAL SAMPLE(N=95)

GENDER RACE

71.639.467.447.480.060.0

55.840.370.567.485.320.0

14.745.312.660.072.665.3

(4)a(15)(6)

(12)(2)

(9.5)

(11)(14)(5)(7)(1)

(16)

(17)(13)(18)(9.5)(3)(8)

FEMALE(N=32)

78.125.868.840.681.365.6

53.143.884.475.087.512.5

12.531.36.3

53.381.371.9

MALE(N=63)

68.346.066.750.879.457.1

57.138.763.563.584.123.8

15.952.415.963.568.361.9

BLACK(N=23)

82.643.569.643.569.634.8

56.545.569.665.282.613.0

8.743.58.7

60.956.552.2

OTHER(N=72)

68.138.066.748.683.368.1

55.638.970.868.186.122.2

16.745.813.959.777.869.4

247Boschi et al

Table 3

COPING AND PSYCHOSOCIAL CHARACTERISTICS OF RESPONDENTS WITH AND WITHOUT COPINGINFORMATION

VARIABLE

CopingActive-CognitiveActive-BehavioralAvoidance

6-Month FunctioningBPRSSANSSAPSGAFBradbum Happiness ScaleQLS Social FunctioningQLS Role FunctioningHamilton Depression Scale

24-Month Social FunctioningBPRSSANSSAPSGAFBradbum Happiness ScaleQLS Social FunctioningQLS Role FunctioningHamilton Depression Scale

WITH COPING INFORMATIONM SD N

3.65 1.403.40 1.322.71 1.20

33.531.661.14

37.863.052.922.089.80

31.651.770.93

39.953.243.072.017.80

9.161.000.02

12.781.481.421.615.42

7.850.890.88

12.201.511.341.584.80

WITHOUT COPING NFORMATIONM SD N

959595

9595959592949191

8081818179818079

33.182.230.84

35.702.442.241.67

10.81

31.522.050.67

35.742.742.472.238.22

9.250.910.937.07.72

0.96.52

5.45

7.880.90.78

8.931.481.281.666.63

3333333327292833

2323232719212223

toms, but found that active-behavioral strategies(e.g., talking with professional) were the mosthelpful.

Gender and RaceThe mean number of coping items endorsed did

not differ by gender (M=9.76 for males, 9.72 forfemales, t=-.08, NS) or by race (M=9.04 forblack, 9.97 for nonblack, t=1.57, NS). In addition,no gender differences appeared in the mean scoreson the three coping style scales: M=3.68 for men,3.59 for women (t=-. 28, NS) for active-cognitive;3.30 and 3.56 (t=.93, NS), respectively, for active-behavioral; and 2.78 and 2.56 (t=-.85, NS), re-spectively, for avoidant strategies. The same pat-tern held for black compared to nonblack respon-dents on the three styles of coping. Nor were anysignificant gender or race differences found in theendorsement of individual coping strategies, senseof control, or degree of stress.

Relationships Among Coping StrategiesCorrelation coefficients among coping styles

can be seen in the first four rows of TABLE 4.Active-cognitive coping strategies were positivelycorrelated with active-behavioral strategies (r=

.42), but neither of these active coping strategieswas significantly related to avoidant coping.

Perceived stress and sense of control were notcorrelated with each other (r=-.04, NS) or with

active behavioral, active coping, avoidant, or totalcoping strategies (r=-.03, -. 08, .09, and -. 02, re-spectively, for perceived stress, and .03, .00, .06,

Table 4PEARSON CORRELATIONS: RELATIONSHIP BETWEEN

COPING AND PSYCHOLOGICAL FUNCTIONING

FACTORCoping

1. Active-Cognitive2. Active-Behavioral3. Avoidance4. Total Strategies

Psychosoc. FunctioningAt 6 Months

BPRSSANSSAPSGAFBHSQLS Functioning:

SocialRole

HDSAt 24 Months

BPRSSANSSAPS

GAFBHSQLS Functioning:

SocialRole

HDS

1 2 3 4

1.00 0.42* 0.03 0.76*1.00 0.01 0.73*

1.00 0.48*1.00

0.21-0.020.18

-0.17-0.26

-0.14-0.110.22

-0.100.00

-0.13-0.20-0.01

0.000.070.04

0.30*0.110.19

-0.13-0.01

-0.14-0.140.25*

0.040.010.09

-0.18-0.05

0.070.030.02

0.010.00

-0.09-0.030.14

0.05-0.050.04

0.060.040.07

-0.020.03

0.00-0.070.04

0.26'0.050.15

-0.17-0.08

-0.13-0.160.27*

0.000.020.01

-0.19-0.02

0.040.020.05

Note. BPRS=Bnef Psychiatric Rating Scale; SANS=Scale for As-sessment of Negative Symptoms; SAPS=Scale for Assessment ofPositive Symptoms; GAF=Global Assessment of Functioning;BHS=Bradburn Happiness Scale; QLS=Quality of Life Scale;HDS=Hamilton Depression Scale.p<.01, two-tailed.

Coping With Psychotic Symptoms in Schizophrenia

and .05, respectively, for sense of control). Norwere coping, stress, or control significantly relatedto the social characteristics or premorbid function-ing variables (analyses not shown).

Finally, respondents were grouped by whetherthey identified an active (cognitive or behavioral)or other (e.g., avoidant) coping strategy as mosthelpful in dealing with their symptoms. A series ofcomparisons, using the variables listed in TABLE 1,yielded no significant differences.

Psychosocial FunctioningAt Six Months

Relationships of the coping variables to psycho-social functioning at the six-month follow-up canbe seen in the middle rows of TABLE 4. After appli-cation of a Bonferonni correction of p<.002(.05/24 comparisons), no significant associationwas found between coping strategies and any six-month functioning variable. There was, however,a trend (p<.01) for active-behavioral and total cop-ing strategies to be positively associated withBPRS scores (r=.30, and .26, respectively). Thenumber of avoidant strategies was not significantlyrelated to any of the psychosocial functioning vari-ables. Finally, perceived stress and sense of con-trol were not statistically associated with any of thesix-month psychosocial functioning variables (anal-yses not shown).

Thus, employment of more and different typesof strategies per se was not related to positive out-comes on the psychosocial functioning measuresat the six-month assessment. In fact, the trend was

just the opposite; the more numerous the copingstrategies respondents endorsed, the more symp-tomatology and depression they experienced.

At Twenty-Four MonthsCoping style and total coping strategies were not

significantly related to functioning at 24-monthfollow-up (see TABLE 4, rows 13-20), and per-ceived stress and sense of control were not associ-ated with the 24-month psychosocial functioningvariables (analyses not shown)

In a series of comparisons, respondents whoidentified an active coping strategy as most helpfuldid better at the 24-month follow-up than did re-spondents who identified other strategies (e.g.,avoidance) or no strategy as most helpful. In par-ticular, they were less symptomatic as measuredby the BPRS (29.4+6.6 vs. 34.2+7.3, t=3.02, p<.01), SANS (1.6+.84 vs. 2.1+.83, t=2.91, p<.01),and SAPS (.7±.8 vs. 1.2+.9, t=2.38, p<.05); scoredhigher on the BHS (3.6±1.3 vs. 2.7±1.8, t-2.25,p<.05); were rated higher on the GAF (42.9±12.9vs. 35.4+9.7, t=-2 .6 7, p,.01) and more proficientin their social functioning (3.3±1.1 vs. 2.8±1.0, t=-2.03, p<.05) and role functioning (2.8±1.9 vs.1 .4+ 1.0, t=-3.7, p<.4 vs. 9.22±4.4, t=2.09,p<.05).

How robust was the positive effect on psychoso-cial functioning of identifying an active copingstrategy as most helpful? This was tested by re-gressing each of the 24-month measures on a)whether an active coping strategy was selected asthe most helpful, and b) the six-month score on thesame measure. Findings from the eight resultant

Table 5OLS REGRESSION EQUATIONS SHOWING PREDICTIVE EFFECT OF SELECTING

ACTIVE COPING STRATEGY AS MOST HELPFUL ON 24-MONTH PSYCHOSOCIAL FUNCTIONING,CONTROLLING FOR 6-MONTH FUNCTIONING

INDEPENDENT UNSTANDARDIZED24-MONTH DEPENDENT VARIABLEBrief Psychiatric Rating (BPRS)

Scale for Assessment of Negative Symptoms (SANS)

Scale for Assessment of Positive Symptoms (SAPS)

Global Assessment of Functioning (GAF)

Bradburn Happiness Scale (BHS)

Social Functioning (SF)

Role Functioning (RF)

Hamilton Depression Scale (HDS)

VARIABLEBPRS-6 months+Active Coping

SANS-6 months+Active Coping

SAPS-6 months+Active Coping

GAF-6 months+Active coping

BHS-6 months+Active Coping

SF-6 months+Active Coping

RF-6 months+Active CopingHDS-6 months+Active Coping

COEFFICIENT (SE)0.42

-4.340.55

-0.220.34

-0.430.147.310.270.810.440.350.271.280.33

-2.15

(0.08)'*'(1.42)-'(0.08)''(0.17)(0.10)***(0.18)'(0.10)(2.80)*(0.12)**(0.36')(0.10)**(0.30)(0.11)**(0.35)''(0.09)'(1.02)

0.280.360.430.450.140.200.030.110.070.130.260.270.110.260.160.21

248

R2 N

74

75

75

77

73

75

72

70

.p.051 .<. 0 1 "*P<00 1

Boschi et al

regression analyses are shown in TABLE 5. As canbe seen, the positive effect held in six of the eightequations. In particular, identifying an active cop-ing strategy as the most helpful was related tolower 24-month BPRS, SAPS, and HDS scores,and to higher 24-month GAFs and role function-ing. Only for the 24-month SANS and social func-tioning was there no significant effect, controllingfor the six-month scores.

DISCUSSIONThe current study's findings on coping with psy-

chosis at a relatively early stage of illness concurwith those of previous reports that individuals withschizophrenia spontaneously employ a wide rangeof coping strategies to deal with their symptoms(Carr, 1988; Lee et al., 1993). Interestingly, themost frequently endorsed strategies were cognitivein nature, while those perceived as most helpfulwere behavioral. In fact, avoidance strategies wereendorsed as most helpful nearly as frequently ascognitive strategies, even though they were theleast frequently used. This result is similar to thefinding by Carter, Mackinnon, and Copolov (1996),who noted that the most frequently reported strate-gies were not those reported as successful.

People with more severe symptoms, as mea-sured by the BPRS, tended to endorse a greaternumber of coping strategies, suggesting that em-ployment of more strategies is a response to symp-toms. However, using more strategies does not perse reduce the distress associated with severe symp-toms, nor does it necessarily lead to better out-comes later.

No significant differences by gender were foundin the use of particular coping strategies. However,the results for gender were in the expected direc-tion; e.g., 84% of women compared to 64% of menreported talking with a professional. This findingis consistent with general population reports ofgender differences in service utilization patterns(Thoits, 1995). However, some studies of the se-verely mentally ill have not found such typicalgender differences. For example, Falloon and Tal-bot (1981) found that women used significantlymore illegal drugs to cope with auditory hallucina-tions than did men, and Carter and colleagues(1996) found that men were less likely than wom-en to ignore voices by trying to focus on some-thing else. Whatever the reason for these unex-pected findings, more research on gender differ-ences is clearly warranted.

Race was not a source of significant differencesin the current study, a finding consistent with thoseof the only previous study examining this aspect(Falloon & Talbot, 1981). Also, the present studyfound no relationship between coping and otherdemographic or background characteristics. Spe-cifically, neither type of coping strategy (cogni-tive, behavioral, or avoidant) nor total number ofstrategies was related to social characteristics, ill-ness onset factors, or other clinical history fea-tures. The most parsimonious explanation for theseresults is that aspects of the disorder itself play amore important role in coping than do demograph-ic or illness onset factors.

Most respondents reported feeling little controland high distress. Control was not, however, corre-lated with any of the outcome measures, nor was itrelated to level of distress. If coping is a responseto symptom severity, these results suggest that useof a coping strategy does not lead to feelings ofcontrol or to less distress. Results also indicatedthat control and distress were not correlated withemployment of either few or many strategies.

It was surprising that only 29% of subjects re-ported low distress, given the global impact of psy-chosis on these individuals' lives. In this regard,Nayani and David (1996) noted that over time, bya process they call "accretion," an individual suf-fering from auditory hallucinations is apt to be-come more involved with the voices, have ex-tended dialogues with them, and describe them inmore intimate detail. The researchers proposedthat, while this process decreases distress, the re-sult is the "ever increasing encroachment of thehallucination into the patient's life resulting inpsychological incapacity" (p. 186). Their findingssuggest that in order to cope, individuals withschizophrenia eventually accept the voices as partof themselves. Other research has supported thisidea. For instance, Farhall and Gehrke (1997)found that patients who took the greatest action tostop the voices had the highest level of perceiveddistress, while Romme et al. (1992) showed thatrespondents they identified as copers interpretedauditory hallucinations as a positive experienceand controlled the voices by "communicating withthem in a selective manner" (p. 102).

Finally, and most importantly, identifying an ac-tive, as opposed to avoidant, coping strategy earlyin the disease process as the most helpful way ofdealing with positive psychotic symptoms was re-lated to a significant improvement in severity of

249

Coping With Psychotic Symptoms in Schizophrenia

symptoms and functioning 18 months later. Thesefindings mesh with those of MacDonald et al.(1998), which found that respondents who felt thatthey coped well with a stressful situation usedmore active coping strategies than did those whofelt that they did not cope well. The better copersalso exhibited fewer negative symptoms.

Limitations of the StudyAmong the limitations to this study is the lack of

data on how often subjects applied each strategyand how effective they felt each one to be. Furtherstudies might examine the fit between certain cop-ing strategies and specific types of symptoms, al-lowing for a better understanding of how individu-als try to cope (Carter et al., 1996). Knowledgeabout how people with severe mental illness cantailor active coping responses to specific stressfulaspects of positive symptoms might lead to moreeffective intervention programs. Although copingmay not reduce the stress associated with psychoticsymptoms, moving people away from ineffectivestrategies, such as drug use or trying to ignoresymptoms, may lead to some improvement in theirphysical and interpersonal lives.

Second, most of the symptomatic respondents inthis sample were prescribed medication, and al-most half were very compliant, suggesting thatmedications should be added to the list of possiblecoping strategies. The role of medication in copingwith positive symptoms remains obscure in thisstudy, however.

Third, the current study posed only one questionabout the general stressfulness of respondents'symptoms. Research with a more detailed focus onthe link between symptoms and distress might giveinsight into the coping process. Studies of generalpopulations indicate that stress varies across differ-ent domains of people's lives (Pearlin & Schooler,1978), and that some coping strategies work betterfor certain types of stressors than for others. Is itsomething about the symptoms themselves that isdistressing, or are they stressful because they makework, interpersonal relationships, and other aspectsof people's lives more difficult? If coping strate-gies work best when they are tailored to the stres-sor in the general population, is the same true forthe severely mentally ill?

Fourth, this study examines coping with positivesymptoms. Other studies have suggested that nega-tive symptoms are also disruptive to social interac-

tions, working, and quality of life. Since most cop-ing studies focus on positive symptoms (Carter etal., 1996; Falloon & Talbot, 1981; Nayani & Da-vid, 1996), future research should include negativesymptoms in the analyses in order to discover ifsuch symptoms modify the relationship betweenpsychotic symptoms and coping. Although the pre-sent study shows no impact of coping on SANSscores at either six or 24 months, accounting fornegative symptoms in the assessment of copingwith positive symptoms may give a fuller-pictureof the coping process.

Finally, further research is needed to investigatewhether coping patterns change over the course ofthis illness. The present study examined coping atone point in time. Other research (Romme et al.,1992) has suggested that coping may evolve, withineffective strategies being dropped over time.However, there is still little information on whatconstitutes successful coping for the severely men-tally ill.

Despite these limitations, the present study in-vestigates coping strategies in a more systematicfashion than has much previous research. Somestudies have asked only about successful strategies,or used open-ended questions (Cohen & Beck,1985; Falloon & Talbot, 1981; Lee et al., 1993;Nayani & David, 1996). Such data collection meth-ods often fail to elicit information on strategies thathave been tried but did not work. The open-endedquestion format is not systematic in querying re-spondents about a particular set of coping strate-gies; although the format allows discussion of abroad range of strategies, we cannot be sure thatthe same set is being examined across all respon-dents. Employing a checklist, as in the presentstudy, may limit the range of coping strategiescited by respondents, but it insures that thoseelicited constitute a standardized set (Carter et al.,1996).

The present study employs a number of outcomemeasures that yield data ranging from symptoms toquality of life. The BPRS, SANS, SAPS, QLS, andHDS are all widely used measures with good valid-ity and reliability. Most studies of coping amongthe severely mentally ill do not cover so wide arange of outcomes.

Clinical ImplicationsNotwithstanding its limitations, the results of the

present study have implications for clinical inter-

250

Boschi et al

ventions with those suffering from a mental illness.First, although the majority of participants spoke tosomeone about their psychotic symptoms, 17% (16of 95) did not, and 30% (28 of 95) did not endorseconsulting a professional as a means of copingwith their psychotic symptoms. Such isolation withtheir symptoms may stem from any of several fac-tors, including stigma, fear of hospitalization, lackof health insurance, and fear of an increase in psy-chotropic medications. Whatever the reason, itposes a serious barrier to prevention and treatment.

Individuals with schizophrenia spontaneouslyemploy a wide range of coping strategies, and in-terventions can build on those that have provedsuccessful. Support groups designed specificallyaround coping with psychotic symptoms will allowgroup members to share their coping experiences,a process that may help them focus on positive, ac-tive strategies when faced with the stress of psy-chotic symptoms. Persaud and Marks (1995) haveshowed that coping enabled drug-resistant schizo-phrenic subjects to gain a sense of control overtheir hallucinations and decreased their sense ofanxiety. While these subjects did not demonstratea decrease in the frequency of auditory hallucina-tions, Persaud and Marks pointed out that theirsense of control over or mastery of the hallucina-tions might be just as important as reducing theiroccurrence.

The results of the present study give reason forsome optimism. Interventions that motivate clientsto engage in management of their symptoms notonly have long-term positive consequences forpsychopathology and social functioning, but arealso compatible with current views that patientsshould be active participants in their own treat-ment. An approach to clinical practice that focuseson clients' abilities to cope with psychotic symp-toms can instill in them a sense of self-efficacy anda belief that they can grow and change despite theirdisability. Research on the positive effects of suc-cessful coping can aid clinicians in their endeavorsto develop programs that help the severely men-tally ill to take better control of their lives.

ReferencesAmerican Psychiatric Association. (1987). Diagnostic and sta-

tistical manual of mental disorders (3rd ed., rev.). Washing-ton DC: Author.

American Psychiatric Association. (1994). Diagnostic and sta-tistical manual of mental disorders (4th ed.). WashingtonDC: Author.

Andreasen, N.C., Amdt, S., Alliger, R., Miller, D., & Flaum,M. (1995). Symptoms of schizophrenia. Archives of GeneralPsychiatry, 52, 341-351.

Bradbum, N.M. (1969). The structure of psychological well-being. Chicago: Aldine.

Breier A, & Strauss, J.S. (1983). Self-control in psychiatric dis-orders. Archives of General Psychiatry, 40, 1141-1145.

Bromet, E.J., Schwartz, J.E., Fennig, S., Geller, L., Jandorf, L.,Kovasznay, B., Lavelle, J., Miller, A., Pato, C., Ram, R., &Rich, C. (1992). The epidemiology of psychosis: SuffolkCounty mental health project. Schizophrenia Bulletin, 18,243-255.

Bromet, E.J., Jandorf, L., Fennig, S., Kovasznay, B., Ram, R.,Tanenberg-Karant, M., & Craig, T. (1996). Suffolk Countymental health project: Demographic, pre-morbid, and clini-cal correlates of 6-month outcome. Psychological Medicine,26, 953-962.

Carr, V. (1988). Patients' techniques for coping with schizo-phrenia: An exploratory study. British Journal of MedicalPsychology, 61, 339-352.

Carter, D.M., Mackinnon, A., & Copolov, D.L. (1996). Pa-tients' strategies for coping with auditory hallucina-tions. Journal of Nervous and Mental Disease, 184,159-164.

Cohen, C.I., & Berk, L.A. (1985). Personal coping styles ofschizophrenic outpatients. Hospital and Community Psychi-atry, 36, 407-410.

Falloon, I.R.H., & Talbot, R.E. (1981). Persistent auditory hal-lucinations: Coping mechanisms and implications for man-agement. Psychological Medicine, 11, 329-339.

Farhall, J., & Gehrke, M. (1997). Coping with hallucinations:Exploring stress and coping framework. British Journal ofClinical Psychology, 36, 259-261

Fennig, S., Craig, T.J., Tanenberg-Karant, M., & Bromet, E.J.(1994). Comparison of facility and research diagnoses infirst-admission psychotic patients. American Journal of Psy-chiatry, 151, 1423-1429.

Gibson, P.J., Lavelle, J., & Bromet, E.J. (1999, June). Buildinga useful pharmacotherapy data set from multiple sources.Poster presented at the annual meeting of the Association ofHealth Services Researchers, Chicago.

Hamilton, M. (1960). A rating scale for depression. Journal ofNeurology and Neurosurgical Psychiatry, 23, 56-62.

Heinrichs, D., Hanlon, T.E., & Carpenter, W. (1984) The qual-ity of life scale: An instrument for rating the schizophrenicdeficit syndrome. Schizophrenia Bulletin, 10, 388-398.

Holahan, C.J., & Moos, R.H. (1987). Personal and contextualdeterminants of coping strategies. Journal of Personalityand Social Psychology, 52, 946-954.

Lee, P.W.H., Lieh-Mak, F., Yu, K.K., & Spinks, J.A. (1993).Coping strategies of schizophrenic patients and their rela-tionship to outcome. British Journal of Psychiatry, 163,177-182.

MacDonald, E.M., Pica, S., McDonald, S., Hayes, R.L., &Baglioni Jr., A.J. (1998). Stress and coping in early psy-chosis. British Journal of Psychiatry, 172 (Suppl. 33),122-127.

McDowell, I., & Newell, C. (1996). Measuring health: A guideto rating scales and questionnaires. New York: Oxford Uni-versity Press.

Moos, R.H., Finney, J., & Cronkite, R. (1990). Alcoholismtreatment: Context, process and outcome. New York: Ox-ford University Press.

National Institute on Drug Abuse. (1988). National household

251

Coping With Psychotic Symptoms in Schizophrenia

survey on drug abuse: Main findings. (DHHS PublicationNo. ADM 88-1586). Washington DC: U.S. GovernmentPrinting Office.

Nayani, T.H., & David, A.S. (1996). The auditory hallucina-tion: A phenomenological survey. Psychological Medicine,26, 177-189.

Pearlin, L.I., & Schooler, C. (1978). The structure of coping.Journal of Health and Social Behavior, 19, 2-21.

Persaud, R., & Marks, 1. (1995). A pilot study of exposure con-trol of chronic auditory hallucinations in schizophrenia.British Journal of Psychiatry, 167, 45-50.

Romme, M.A.J., & Escher, A. (1989). Hearing voices. Schizo-phrenia Bulletin, 15, 209-216.

Romme, M.A.J., Honig, A., Noorthoorn, E.O., & Escher,A. (1992). Coping with hearing voices: An emancipatory

approach. British Journal of Psychiatry, 161, 99-103.Spitzer, R.L., Gibbon, M., Williams, J.B.W., & Endicott, J.

(1996). Global assessment of functioning (GAF) scale. InL.I. Sederer & B. Dickey (Eds.), Outcomes assessment inclinicalpractice (pp. 76-78). Baltimore: Williams & Wil-liams.

Spitzer, R.L., Williams, J.B.W., Gibbon, M., & First, M.B.(1992). The structured clinical interview for DSM-III-R(SCID). Archives of General Psychiatry, 49, 624-629.

Thoits, P. (1995). Stress, coping and social support processes:Where are we? What next? Journal of Health and SocialBehavior, 53-79.

Woerner, M.G., Mannuzza, S., & Kane, J.M. (1988). Anchor-ing the BPRS: An aid to improved reliability. Psychophar-macology Bulletin, 24, 113-121.

For reprints: Evelyn Bromet, Ph.D., Department of Psychiatry and Behavioral Science, Putnam Hall-South Campus, StateUniversity of New York, Stony Brook, NY 11794-8790 [E-mail: [email protected]]

252