2 Method go

2.1 Sample Description go

2.2 Study, Design and Procedure go
2.3 Instrumentation go

2.1 Sample Description 

Participants of the investigation in hand were initially 381 coronary bypass surgery patients who responded to the presurgery questionnaire at the first wave of data collection. For 122 out of the 381 patients at point in time 1 there was also a close relative who responded to the relatives’ questionnaire.

The average age of the patients was 59.84 years (SD = 10.33), the youngest being 25 years old, the oldest 82 years. The sample consists of 302 male heart patients, and 79 female heart patients.

297 patients reported to be married, 28 were divorced, seven had split up with their partner, 29 were widowed, and 18 were singles (two missing values for family status). Out of 345 patients 51 stated to be living alone, 294 stated to be living together with a partner (36 missing values).

Not actively in the workforce (retired, unemployed, or homemakers) were 61.7% (n = 235) of the patients, 34.9% (n = 133) reported to be working (13 missing values). Of those who had a job 27.6% (n = 105) were blue collar workers, 55.9% (n = 213) reported to be white collar workers (23 missing values).

The average age of the relatives was 51.42 years (SD = 12.09), ranging from 25 years to 75 years (eight missing values for age). The subsample consists of 18 male relatives, and 103 females (one missing value for gender). Both statements pertain to point in time 1. At point in time 3, the relatives’ longitudinal subsample consists of 51 persons. The mean age was 53.92 years (SD = 11.91), ranging from a minimum of 25 years to a maximum of 71 years. The subsample contains eight male persons, and 43 female persons at follow-up.

 The longitudinal subsample at point in time 2 contains 247 patients who had completed the postsurgery interview in addition. The attrition rate was 35.5% for the following reasons: 19.4% (n = 74) were transferred to different hospitals early after surgery, 5.8% (n = 22) passed away, 5.5% (n = 21) did not undergo surgery, and 4.5% (n = 17) were unwilling to be interviewed.

Within this longitudinal sample there were 193 men with a mean age of 59.1 years (SD = 10), and 55 women with a mean age of 57.4 years (SD = 11.7). Coronary artery bypass surgery was performed on 152 of the 193 men, and on 26 of the 55 women.

The others underwent different kinds of heart surgery, with most patients having been scheduled for cardiac valve substitution, heart transplantation, removal of heart tumors, or aneurysm resection. More men (n = 159,

82.46%) than women (n = 19, 34.5%) had a life partner. A myocardial infarction had been experienced by 35% of the patients, two infarcts by 6.4%. Moreover, 48.1% of the males were retired or jobless, compared to 26.1% of the women.

At point in time 3 the patient sample still consists of n = 209 persons who completed the week books, and the final questionnaire.

At the first stage of data analysis, a heterogeneous sample of 122 relatives was available. They were 18 men, and 102 women who responded at the first wave. Most of the relatives were spouses (13 husbands, and 95 wives). Also, there were two sons, and six daughters. Moreover, we find two other male family members, and one male and one female ‘friend’.

Since the patients were asked to give the questionnaire to a person with whom they are in close contact, for example, their wife/husband, we will speak of relatives despite the two ‘friends’, because it can be assumed that the patients’ relationship with these persons is very close.

2.1.2 Description of Characteristics of the Cross-Sectional Subsample of Social Network Members
Means and standard deviations for a variety of characteristics are given in Table 2.1. The average age for the men is 53 years and the average age for women is 51. There is no significant age difference between those two groups. As is the case for the majority of variables.

There are a few exceptions though. Women report higher social support than men do. At the same time they report less loneliness than men. Moreover, their scores are significantly higher on two of the four Self Regulation Competence (SRC) subscales (on Conflict Avoidance, and on Motivation Control). For ‘Indecisiveness’ the difference between female and male relatives is still significant on a 8% level, explaining 3% of the variation.

Table 2.1: Relatives’ Characteristics at Time 1

Treating Physical Symptoms in accordance to general findings, women are reporting more symptoms (on a significant level much above chance, p = .15, h ² =.02), and significantly more headaches, and more troubled sleep than males do. The opposite pattern arises for males.

Thus, a consistent pattern is found: Female relatives feel more supported. Presumably this is the reason for them to feel less lonely (and almost significantly more optimistic, p = .07, h ² = .03) than males. Furthermore, females show more stamina, and they are willing to take action. At the same time, they report more symptoms which might be classified as rather psychosomatic.

2.1.3 Differences between the Patient Sample and the Subsample of Relatives
With respect to a comparison between the whole sample of patients (N = 381), and the subsample of relatives (n = 122) there were some differences found on the major variables. The most prominent variable among them is Physical Symptoms, on which astonishingly relatives are much higher. A significant difference is also found for age (average for patients is 60 years, for relatives it is 52 years). Since most of the relatives are females (n = 103) this result does not come as a surprise.

The remaining significant mean differences between relatives and patients are pertaining to loneliness, optimism, as well as to two Self Regulation Competence (SRC) subscales, namely ‘Indecisiveness’, and ‘Motivation Control’. All results are reported in detail in Chapter 3.

2.2 Study, Design and Procedure
The study was set up as a longitudinal design with three measurement points in time, once before heart surgery, about one week afterwards, and again three month respectively half a year later. Patients were contacted upon arrival at the heart surgery ward of the Charité Hospital Berlin, and were asked to participate in the study. They were briefed very generally about the research, the purpose of which was declared as an "investigation of the effects of severe disease and surgery on patients’ quality of life". They were assured that the data would be computerized anonymously, and that participation was voluntary. They received a questionnaire to be returned in a sealed envelope as soon as possible, but surely before surgery, into a box that was available on the counter of the ward (Wave 1 of data collection). At the same time, they were asked if by chance a close relative were visiting them before surgery, and if they might be willing to receive a second less voluminous questionnaire for their relative or spouse. This one was also returned into that box, usually at the same time, and in a sealed envelope, too.

Patients were approached again not earlier then five days after surgery for an interview (Wave 2). In those cases where patients were unable to be interviewed (e.g., due to poor physical condition), further attempts to obtain interview data were made in the subsequent days until discharge from hospital. The post-surgery interview took about half an hour, and included questions about physical and mental well-being, and activities such as sitting up in bed, or walking in the hall.

At the end of the interview patients were asked to further participate by accepting to receive a book for weekly report (15 minutes per week) three month after their discharge from hospital. Patients were to report for 10 weeks about their physical and emotional well-being, respectively their adaptation process.

Eventually, patients received a questionnaire, and a self-addressed stamped envelope by mail half a year later. The questionnaire was designed to assess self-reported quality of life (Wave 3).

In case that a patient’s relative had participated in the investigation previously, there was also a second questionnaire included for the patient’s spouse or relative at this point in time 3.

2.3 Instrumentation
A variety of instruments were used in this investigation. Only those will be mentioned of which results are reported.

To put them in a certain order, first measures used only for patients will be listed, followed by measures that were used both for patients, and for their relatives.

2.3.1 Measures for Patients

2.3.1.1 Trier Coping with Illness Scales
At the first point in time the patients’ questionnaire contained the Trier Coping with Illness Scales by Klauer and Filipp (1993). The five subscales explore the adaptivity of different coping styles.

All items could be answered on a five-point scale ranging from "never" to "very often".

The first subscale is ‘Threat Minimization’. It consists of eight items like "I told myself that I am just going through hard times, and that there can be a lucky future ahead.". Klauer, and Filipp found an a of .75.

The second subscale is called ‘Rumination’, a = .80. A typical item is, e.g., "I was absorbed in daydreams.". The scales contains nine items.

The third one is ‘Search of Information and of Exchange of Experiences’ with an a of .83. It consists of eight items like "I exchanged experiences with other patients about how to cope with the illness.".

The next scale is called ‘Search of Support in Religion’. Here the a = .77. The scale consists of three items only, like "I prayed for strength to solve my problems.".

The last of the five subscales is ‘Search of Social Integration’ with an a of .76. The scales contains nine items such as "I visited other persons or invited them.".

2.3.1.2 Coping with Everyday Problems
The next measure only applied to patients is called Coping with Everyday Problems by Bullinger, Kirchberger, and von Steinbuechel (1993). It consists of six subscales. They are introduced in the questionnaire with the sentence:"In the past couple of weeks were you able to...", followed by 42 questions referring to one’s coping. Answers are possible on a five-point scale ranging from "not at all" to "very well".

The first scale is ‘Every Day Life’. The authors found an a of .89. All in all it contains nine items. One of its typical items is "...manage your household chores?".

Next there is the subscale ‘Body’, with an a = .64, and nine items in total. There are items such as "...to sleep in a way that you feel thoroughly refreshed in the morning?" to be found.

The third subscale is called ‘Joy of Living’. It has only three items, a = .78. Items run like "...enjoy life?", or "...confront future confidently?".

Fourth, there is a subscale called ‘Satisfaction with Medical Care’, with an a of .72. Also this scale consists of just three items. There are items such as "...feel satisfied with the medical care?".

Next is the subscale ‘Psyche’. It contains nine items, a = .89. Here items like "...care for your appearance?", or "...like yourself/accept yourself?" are to be found.

Last is the subscale ‘Social Integration’. It contains nine items such as "...participate in family life?". Alpha was found by the authors to be .81.

2.3.1.3 Munich List of Quality of Life Dimensions
The Munich list of Quality of Life Dimensions by Heinisch, Ludwig, and Bullinger (1991) contains two subscales. In the first one are 17 items asking for the extent to which someone was satisfied with certain aspects of his or her quality of life. It is introduced by the sentence:"I was satisfied with my...". A typical item of the subscale ‘Satisfaction’ is "...physical condition", to be answered on a five-point scale ranging from "very much satisfied" to "very dissatisfied". For this subscales the authors found an a of .87.

The second subscale ‘Importance’ contains the same 17 items. But this time it asks for the importance of each of the aspects of quality of life. It is introduced with the sentence:"For me my...". Items run like "...physical condition.", again to be answered on a five-point scale ranging this time from "very important" to "very unimportant". Alpha was found to be .90.

The instrument was used at point in time 3, only.

2.3.1.4 Impact of Physical Condition on Everyday Functioning
The scale consists of three items, alpha was found to be .80 in the present investigation. The items are introduced by the sentence:"How strongly did you feel in the last week...". An item runs like "...impaired by your physical condition in your everyday life?". Answers had to be given on a five-point scale ranging from "not at all" to "very much".

2.3.1.5 Health Locus of Control Scale
The last measure that was only used in patients` questionnaires is the Health Locus of Control Scale by Filipp, and Ferring (1993). It is devided into three subscales, and was used at point in time 1. It is to be answered on a six-point scale ranging from "completely wrong" to "completely right".

The first subscale ‘External-Powerful Others’ contains five items, a was found to be .62. It refers to powerful others as the source for increasing one’s well-being. There are items such as "Actually I can only do what doctors recommend to me." to be found.

Also in the second subscale ‘External-Chance’ the locus of control is external, this time referring to chance. A typical item is "When I am feeling bad I can just wait, and hope that I will soon be better again.". Alpha was = .77 for ‘External-Chance’.

The last one of these three subscales is called ‘Internal’, containing ten items with an a of .86. Items run like "I can do a lot of things myself to get well again.", or "Having the right attitude one can even master serious illness.".

2.3.2 Instruments Used for Both Patients and Their Relatives

2.3.2.1 Profile of Mood States (POMS)
The Profile of Mood States (POMS) by Bullinger, Heinisch, Ludwig, and Geier (1990) was used at all points in time for both patients and relatives. The items are adjectives that pertain to emotions. They are introduced by the sentence:"I felt...". Subjects have to mark their answers on a five-point scale ranging from "not at all" to "very strongly". It contains four subscales. The last item of the scale, namely ‘decisive’, was taken as a single item in addition.

The first subscale is ‘fatigue’, consisting of seven items. Bullinger et al. found an a of .91. Typical items are "...tired", or "...exhausted".

The second one is ‘displeasure’. This subscale also contains seven items, a = .86. Typical adjectives are "...irritated", or "...angry".

The third subscale is called ‘sadness’, a here is .93. It consists of fourteen adjectives. Typical ones run like "...discouraged", or "...helpless".

The last Profile of Mood States (POMS) subscale is called ‘vigour’, and consists again of seven items. Alpha is found to be .90. Typical adjectives are "...gay", or "...vigorous".

2.3.2.2 List of Physical Symptoms
The List of Physical Symptoms by von Steinbüchel, and Haekel (1991) contains eighteen items. The authors found an a of .81. The instrument was developed for patients suffering from hypertension. It is introduced with the sentence:"Did you suffer from the following symptoms during the last week?". Answers are to be given on a five-point scale ranging from "not at all" to "very strongly". Typical symptoms are "everything went black", or "troubled sleep".

In the present investigation the list was slightly altered to adapt it to the needs of bypass surgery patients. The alterations were done in co-operation with the medical doctors at the Berlin Charité Hospital. In addition, there were changes from point in time to point in time, e.g. asking for "pain from operation wound" only at the second point in time.

Obviously the List of Physical Symptoms also has to be different for relatives. For relatives it contained more general symptoms of exhaustion, or overstrain such as "backache", or "stitches, pain, or twinges in your chest".

For patients the List of Physical Symptoms was present at all three points in time, and also for relatives at both points in time.

2.3.2.3 Self Regulation Competence (SRC)
Another predominant instrument is the Self Regulation Competence (SRC) Scale by Schröder (1993). It was applied to both patients and relatives at all points in time with the only exception of the patients’ interview data (point in time 2).

In general, this scale with all its subscales is pertaining to peoples’ ability to regulate themselves in the field of decision making, motivation, goal-directed behaviour, and coping with forthcoming difficulties and conflicts.

In the present investigation 38 items of three subscales were used. The items were to be answered on a four-point scale ranging from "not at all true" to "absolutely true".

The first subscale is called ‘Goal-Directed SRC’. Schröder found an a of .88. There is a further distinction within ‘Goal-directed SRC’ into two subscales, namely ‘Indecisiveness’, and ’Conflict Avoidance’, referring to the degree to which people stick to their decisions, and to which degree they are able to stand conflict, respectively, while behaving in a goal-directed manner. These two subscales together consist of 16 items. Typical items run like "I often find it difficult to decide in favour of something." for ‘Indecisiveness’, and like "In case of problems I take the initiative." for ‘Conflict Avoidance’, respectively.

Also the second subscale, ‘Coping-related SRC’, shows a further division into subscales. The one used in this investigation is called ‘Readiness to Feel Resigned’, pertaining to the way of coping with difficulties by not giving up easily. This subscale contains twelve items, a is found to be .91. A typical item, e.g., is "When something is not easy to do I start worrying about my ability to do it at all.".

The last subscale is ‘Action Control’. From this subscale here a choice of another ten items is summarized as ‘Motivation Control’. A typical item of that subscale is "I always start to work on important, and difficult tasks as soon as possible.". Alpha is found to equal .76.

2.3.2.4 Social Support Scale (SSS)
In the study in hand social support was measured by the Social Support Scale (SSS) by Donald, and Ware (1984). Also this instrument was used at all points in time for both groups, except point in time 2 for patients (interview).

The scale consists of 19 items. Donald and Ware found a to be .96. It is introduced by the sentence:"Is there somebody who...", followed by an items such as "...embraces you", or "...helps you in case you are bedridden.". The answers come on a five-point scale ranging from "never" to "always".

The factor analysis of the German version by Kirchberger (1991) showed three dimensions: ‘Cognitive Support’, ‘Emotional Support’, and ‘Tangible Support’, holding for 74,4% of the observed variance.

2.3.2.5 Measures of Traits

2.3.2.5.1 Perceived Self-Efficacy
All items concerning traits like self-efficacy, optimism, loneliness, and anger were mixed within a 38 item block, introduced by the sentence:"Here your attitudes and emotions are concerned. Please, indicate to what extend the following statements are true for you.".

The perceived self-efficacy of patients and relatives was measured by a scale called General Self-Efficacy by Schwarzer, and Jerusalem (1994). The authors found their ten item scale to have an a of .82. A typical items runs like "I have a solution to every problem.", or "Whatever may happen, I will find my way.". These items are to be answered on a four-point scale ranging from "not true" to "exactly true".

2.3.2.5.2 Optimism
The ‘Dispositional Optimism Scale’ by Scheier, and Carver (1985) was used. It consists of eight items. Alpha was found to equal .61. The answering pattern was the same like mentioned above, items were answered on a four-point scale ranging from "not true" to "exactly true".

In the present investigation there were found to be two subscales with four items each, called ‘Optimism’, and ‘Pessimism’, respectively (see Wiethoff, 1996). Typical items are "I always see the positive aspects of things.", and "I almost never expect things to end well in my favour.", respectively.

2.3.2.5.3 Loneliness
Loneliness was measured with a choice of ten items taken from the ULCA-LS University of California at Los Angeles Loneliness-Scale by Russell et al. (1982), items selected by Stephan and Faeth (1989).

For the selected items the authors found an a of .82. The answering pattern is the same, with items to be answered on a four-point scale ranging from "not true" to "exactly true". A typical item e.g. is "I know people who are close to me.".

2.3.2.5.4 Anger
Data about anger were collected with the Trait-Anger Scale from the STAXI State Trait Anger Expression Inventory by Spielberger et al. (1988), German version by Schwenkmezger, Hodapp and Spielberger (1992). The authors found an a of .85 for their ten item scale. Typical items are "I become upset very easily.", or "It makes me angry if me of all people is corrected.". The answering pattern again like above, items are to be answered on a four-point scale ranging from "not true" to "exactly true".

2.3.2.6 Quality of Life
The last instrument used for both patients and relatives to be mentioned is a single item concerning the Quality of Life. It was presented at point in time 3 for patients, and at both points in time for relatives.

The item runs as follows "Nowadays quality of life is often discussed. How would you evaluate the quality of your life for the last week?". Answers went to a five-point scale ranging from "miserable" to "splendid".

2.3.3 Stability Coefficients for the Major Instruments
Re-test coefficients for patients’ and for relatives’ variables were computed for measures of point in time 1 with measures for point in time 3.

Only the major instruments were chosen for this procedure.

2.3.3.1 Stability Coefficients for Patients’ Scales
The re-test coefficients for patients’ variables are all continuously highly significant (see Table 2.3.4.1.1 ), ranging from r = .33^** to r = .81^**. 

Table 2.3.4.1.1: Stability Coefficients for Patients’ Instruments 

 Note. * - Signif. LE ,05 ** - Signif. LE ,01 (2-tailed)

 The highest scores are to be found for social support (r = .81^**), anger (r = .73^**), self-efficacy (r = .67^**), and for all of the three Self Regulation Competence (SRC) scales (from r = .63^** to r = .76^**).

The lowest coefficient is found for the single item ‘decisiveness’ that belongs to the Profile of Mood States (POMS) subscale ‘vigour’. For the complete subscale the re-test coefficient shows to be much higher, namely r = .52^**.

A noteworthy result is the rather high re-test coefficient for patients’ Physical Symptoms with r = .47^**. So patients suffering from many symptoms at point in time 1 still tend to suffer from many symptoms at follow-up six months later.

2.3.4.2 Stability Coefficients for Relatives’ Scales
The re-test coefficients for relatives' variables show a less prevailing pattern, although most of them are highly significant, too (see Table 2.3.4.2.1). The highest is loneliness with r = .79^**, followed by self-efficacy, and social support (r = .75^**, and r = .73^**, respectively).

Table 2.3.4.2.1: Stability Coefficients for Relatives’ Instruments

Note. * - Signif. LE ,05 ** - Signif. LE ,01 (2-tailed)

But a re-test coefficient of r = .75^** is also found for the Profile of Mood States (POMS) subscale ‘vigour’. The other subscales, however, show a less continuous pattern. For the single item ‘decisiveness’, and for ‘fatigue’ there are found to be highly significant correlations of r = .57^**, and r = .48^**, whereas the subscales ‘sadness’, and ‘displeasure’ just show moderate associations between both points in time (r = .30^*, and r = .09, respectively).

Actually, Profile of Mood States (POMS) ‘sadness’ as well as ‘displeasure’ at point in time 3 show high associations to relatives’ ‘fatigue’, and ‘decisiveness’ at point in time 1.

All relatives’ Self Regulation Competence (SRC) subscales show very high re-test coefficients. This is almost the same result for patients like for relatives.

Worth mentioning also here is a very strong association between relatives’ Physical Symptoms at point in time 1 and point in time 3, namely r = .74^**. This correlation is much higher even than the same re-test coefficient for patients. That is to say, that relatives are extremely stable in their physical symptoms over a six months period.

This finding seems to be reasonable since patients should recover to a certain extent within this time period so that it is to be assumed that their symptoms should not be as stable as their spouses’ symptoms. Much more so, because the List of Physical Symptoms differs from patients to relatives, as mentioned above. Relatives’ symptoms refer more to being exhausted in general. Living together with an recovering heart patient should therefore be mirrored in a certain stability of their symptoms.

© by Gerdamarie Schmitz