PSY C H O LO G Y , H EAL T H & M ED IC IN E , VO L . 4 , N O . 4 , 1999
SH O RT REP O R T
The relationship between control beliefs and self-reported adherence in adults with cystic ® brosis L. B. M YERS 1 & F. M YERS 2 D epartment of Psychiatry & Behavioural Sciences, University College London & 2D epartm ent of Psychology, Royal H olloway, U niversity of London , UK 1
A bstract T his study investigated the relationship between locus of control and self-reported adherence in 31 adults with cystic ® brosis. Participants com pleted the M ultidimensional H ealth Locus of Control Scale, Form C (M H LC -C; W allston et al., 1994) and an adherence questionnaire. The two `powerful others’ sub-scales of the M HLC-C, `doctors’ and `others’ , were signi® cantly correlated with overall self-reported adherence. M ultiple regression indicated that the `doctors’ sub-scale explained 35% of the variance in overall adherence. It is concluded that an external locus of control for powerful others, especially doctors, m ay prom ote greater adherence.
C ystic ® brosis (CF) is a serious genetic disorder which requires com plex daily life-long com m itm ent to m edical regim ens which are necessary to both m inim ize and delay the effects of th e disease process (Harris & Super, 1995). The m ain pathophysiology of CF is observed in the lungs and digestive system and treatm ent includes replacem ent of pancreatic enzym es and nutritional supplem ent, antibiotics and chest physiotherapy (Glasscoe, 1997). The advent of the C F centres, early diagnosis and improved treatm ents have led to the prediction that 50% of patients with CF will reach adulthood (Glasscoe, 1997). H ow ever, whatever th e advances in treatm ent, an important factor is how m any treatm ents patients adhere to and what factors in¯ uence such adherence. Non-adherence to m edical treatm ent is a widespread problem (see M yers & M idence, 1998), with adherence to long-term regim ens being estim ated at around 50% (Sackett & Snow , 1979). Research on C F has indicated that adherence is a problem in both children and adults (Abbott et al., 1994; for a recent review see Bryon, 1998). A search for factors which are associated with poor adherence in CF has been inconclusive. H ow ever, in a recent study of adults with CF, Abbott et al. (1996) m easured health locus of control (HLC ) with the generic M ultidimensional H ealth Locus of C ontrol Scale (M H LC ; W allston et al., 1978), the m ost com m only used m easure of H LC. The M HLC recognizes different control beliefs which are m easured by three separate scales: an internal scale, and A ddress for correspondence: Dr Lynn B. M yers, Department of Psychiatry & Behavioural Sciences, R oyal Free and University College Medical School, University C ollege London, W olfson Building , 48 R iding House Street, L ondon W 1N 8AA, U.K. Tel: 1 44 0171 504 9464; Fax: 1 44 0171 323 1459; E-mail
[email protected] ISSN 1354-8506 print /ISSN 1465-3966 online/99/040387-05
Ó
1999 Taylor & Francis Ltd
388
L . B . M YER S & F . M YERS
tw o external scales, chance and powerful others. Abbo tt et al. noted that certain adherence behaviours (physiotherapy, pancreatic enzym e and vitam in therapies) were related to having an external locus of control, speci® cally to powerful others. The value of the H LC construct as a predictor of health behaviour has been questioned as m easures of general HLC have been shown to be fairly weak predictors of health behaviours (see Norm an & Bennett, 1996). O ne of the m ain criticism s has been the usefulness of using generic m easures of HLC , as it is likely that individuals m ay hold different control beliefs for different behaviours and that speci® c m easures m ay be better predictors of speci® c behaviours (Norm an & Bennett, 1996). Recently, the use of disease-speci® c m easures for assessing H LC has been found to improve the utility of HLC in explaining m edication-related behaviour (e.g. Johnson et al., 1993; W allston et al., 1991). The current study was designed as a prelim inary study to investigate whether adherence was related to the powerful others scale of a disease-speci® c health locus of control scale in adults with CF.
M ethod Participants Participants were patients over the age of 18 years attending a C F centre in the U K. Q uestionnaires were posted to 60 patients with C F, which resulted in a resp onse rate of 51%. Thirty-one participants, 13 fem ales and 18 m ales, responded to the questionnaire (mean age 5 28.05, SD 5 6.67).
M easures M ultidimensional H ealth Locus of Control Scale, Form C (M HLC-C; W allston et al. 1994). This is an 18-item condition-speci® c scale where an existing health/m edical condition is used to m easure health locus of control. There are four sub-scales: internal (six item s), chance (six item s), and pow erful others, which has tw o independent sub-scales, doctors (three item s) and others (three item s). Answ ers are rated on a six-point scale from strongly disagree (1) to strongly agree (6). High scores on each sub-scale indicate high control.
CF adherence questionnaire. This consisted of nine item s relating to key treatm ent areas in C F derived from the m ost com m only prescribed treatm ents and other C F adherence m easures (Abbott et al., 1994; C zajkowski & Koocher, 1987). It covered the m ain aspects of treatm ent: high calorie supplem ents, pancreatic enzym es, taking antibiotics/m edication, physiotherapy/regular exercise, vitam in supplem ents, nebulizer, attending appointm ents (see Table 1). Answers were on a seven-point scale from strongly disagree (1) to strongly agree (7). A m idpoint (4) was labelled average scale to give a point from which low and high adherence could be estim ated. Tw o item s were negatively worded (enzym es/exercise) and were scored in reverse. A high score indicated high self-reported adherence.
D em ographic variables.
Participants were asked for their age and gender.
AD H ERE N C E AN D C YS TIC FIB RO SIS
389
Table 1. M eans (and standard deviations ) for self-reporte d adherenc e item s Q uestion
Mean (SD)
1 often forget to take my enzymes with meals Taking vitamin supplements is part of my daily routine
6.34 (1.82) 6.13 (1.74)
I rarely miss an outpatient appointment 1 always take antibiotics prescribed by my doctor I use my nebulizer at least daily 1 seldom run out of my medication
6.10 5.61 5.10 4.91
R egular exercise is not part of my lifestyle R egular high calorie meals are part of my everyday life Physiotherapy is part of my daily routine
4.41 (1.87) 4.06 (2.00) 3.00 (2.59)
(1.56) (1.73) (2.44) (2.10)
R esults Individual adherence m easures M eans and standard deviations of adherence behaviours can be found in Table 1. To investigate the different ratings of the various adherence behaviours, th e neutral m idpoint of 4 (i.e. average) was taken to indicate neither adherent nor non-adherent behaviour. Adherence scores signi® cantly above the m idpoint were called `h igh adherence behaviours’ and adherence scores signi® cantly below the m idpoint were called `low adherence behaviours’ . Each adherence score was com pared with the m idpoint using one sam ple t-tests. R esults indicated th at som e adherence behaviours were signi® cantly higher than average and were therefore considered high adherence behaviours (antibiotics, t (30) 5 5.20, p , 0.001, m edication, t (30) 5 2.39, p , 0.05, enzym es, t (30) 5 10.12., p , 0.001, vitam ins, t (30) 5 6.79, p , 0.001, nebulizer, t (30) 5 2.51, p , 0.05, and outpatient appointm ents, t (30) 5 7.39, p , 0.001). H igh calorie m eals and exercise did not signi® cantly differ from th e average (t (30) 5 0.18, ns, t (30) 5 1.25, ns, respectively). Physiotherapy was signi® cantly lower than the average (t (30) 5 2 2.15, p , 0.05) and was considered a low adherence behaviour. The relationship between control and overall self-reported adherence For the M H LC -C scales, internal was signi® cantly negatively correlated with chance (r 5 2 0.52, p , 0.01); others was signi® cantly positively correlated with doctors (r 5 0.46, p , 0.01) and negatively correlated with chance (r 5 2 0.53, p , 0.01). Doctors was negatively correlated with chance (r 5 0.34, p , 0.05). C ronbach’ s alpha for the adherence scale was 0.71, which was considered satisfactory, therefore the item s in the scale were analyzed together. Initially, the overall m ean adherence score was correlated with the different scales of the M HLC -C . A s adherence has been associated with age and gender (Czajkowski & Koocher, 1987), age and gender were partialled out in subsequent analyses (see Table 2). Adherence was signi® cantly correlated with the tw o powerful others scales: doctors and others. W hen both age and gender were partialled out the sam e pattern was noted. Adherence was not signi® cantly correlated with internal or chance. To investigate whether the tw o powerful others scales were predictive of self-reported adherence, a m ultiple regression was perform ed, with overall self-reported adherence as the dependent variable and the tw o powerful others scales, doctors and others, as independent variables. This analysis indicated that doctors alone explained 35% of the variance in self-reported adherence (t 5 3.09, p , 0.001, adjusted R square 5 0.35). The form ula given
390
L . B . M YER S & F . M YERS
Table 2. Pearson’ s correlation s between MH LC-C scales and adherenc e
Adherence Chance Doctors Internal Others *p ,
0.05, **p ,
2
0.33 0.61*** 0.13 0.45* 0.01, ***p ,
A dherence: partialling out age
2
0.35 0.57** 0.31 0.44*
Adherence: partialling out gender
2
0.33 0.61*** 0.14 0.45*
0.001.
by Cohen (1988) was used to calculate the power of the m ultiple regression. Setting the acceptable Type 1 error at a conservative ® gure of 0.01, the power for the m ultiple regression was 0.81.
D iscussion The current study replicated and extended the ® ndings of A bbott et al. (1996) using a disease-speci® c locus of control m easure. O verall adherence to treatm ent in adults with C F was signi® cantly correlated with th e two powerful others scales, doctors and others. This indicated th at go od adherence was related to external control, especially the doctors subscale, which predicted over a third of th e variance in self-reported adherence. It should be noted that in th e M HLC -C there is no speci® c health professional m entioned, apart from `doctor’ , and that the doctor and others sub-scales were signi® cantly positively correlated. In future studies, it should be investigated whether adherence is speci® cally related to the control of other health professionals involved in m anaging C F, such as a physiotherapist or a specialist nurse. This study highlighted that physiotherapy adherence seem s to be a particular issue in th is group of C F sufferers. This is sim ilar to previous studies (e.g. Abbott et al., 1994; Passero et al., 1981). A s physiotherapy is a vital treatm ent in prolonging life, further studies sh ould focus speci® cally on designing interventions to improve physiotherapy. In conclusion, this is a prelim inary study wh ich has investigated adults in one C F centre. This study needs to be extended to other C F centres, to see whether external control by powerful others, especially doctors and other health professionals, is related to better adherence behaviours. If ® ndings from such studies show a sim ilar pattern of results, interventions could be developed possibly based on improving com m unication between the C F patient and the doctor or other health professional, to help to facilitate an external locus of control, speci® cally for C F behaviours.
R eferences A BBOTT , J., D ODD , M ., B ILTON , D. & W EBB , A . K. (1994). Treatment compliance in adults with cystic ® brosis . Thorax , 49, 115± 120. A BBOTT , J., D ODD . & W EBB , A. K. (1996). Health perceptions and treatment adherence in adults with cystic ® brosis . Thorax , 51, 1233± 1238. B RYON , M. (1998). A dherence to treatment in children . In: L. B. M YERS & K. M IDENCE (Eds), Adherence to treatmen t in medical condition s (pp . 161± 190). H arwood Academic Publisher s. C OHEN , J. (1988). Statistica l power analysis for the behaviora l sciences, 2nd edition. Hillsdale, NJ: Lawrence Erlbau m. C ZAJKOWSKI , D. R. & K OOC HER , G. P. (1987). Medical compliance and coping with cystic ® brosis. Journal of Child Psychology and Psychiatry, 28, 311± 319.
AD H ERE N C E AN D C YS TIC FIB RO SIS
391
G LASSCOE, C. A . (1997). Cystic ® brosis . In: A . B AU M , S. N EW MAN , J. W EINMAN , R . W EST & C. M C M ANUS , The Cambridg e handboo k of psycholog y, health & medicin e (pp. 424± 426). Cambridge: Cambridg e University Press. H ARRIS , A. & S UPER , M. (1995). Cystic ® brosis. The facts, 3rd edition. O xford: Oxford University Press. J OHNSON , L. R ., M AGNANI, B., C HAN , V. & F ERRANTE , F. M. (1993). M odi® ers of patient-controlled analgesia ef® cacy: I. Locus of control. Pain , 39 , 17± 22. M YERS , L. B. & M IDENCE , K. (E ds) (1998). Adherence to treatment in medical condition s. C har: H arwood Academic Publishe rs. N ORM AN , P. & B ENNETT , P. (1996). H ealth locus of control. In: M . C ONNER & P. N ORM AN (Eds), Predicting healt h behaviour (pp. 62± 94). Buckingham: Open University Press. S AC KETT , D. L. & S NOW , J. C. (1979). The magnitude of compliance and noncompliance. In: R. B. H AYNES , D . W . T AYLOR & D. L. S ACKETT (Eds), Complianc e in health care. Baltimore: Johns Hopkins University Press. W ALLSTON , K. A . & W ALLSTON , B. S. (1980). H ealth locus of control scales. In: H . L EFCOURT (Ed.), Advances and innovation s in locus of control research (pp. 198± 234). New York: Academic Press. W ALLSTON , K. A., W ALLSTON , B. S. & D EV ELLIS , R . (1978). Development of the multidimensional health locus of control (M HLC ) scales. Health Education Monograph s, 6, 160± 170. W ALLSTON , K. A ., S M ITH , R. P., K ING , J. E. & S MITH , M . S. (1991). D esire for control and choice of antiemetic treatment for cancer chemotherapy. Western Journal of Nursin g Research, 13, 12± 29. W ALLSTON , K. A., S TEIN , M. J. & S M ITH , C. A. (1994). Form C of the MH LC scales: a condition-speci® c measure of locus of control. Journa l of Personalit y Assessment, 63, 534± 553.
