The relationship between expectations and satisfaction: a qualitative study of patients' experiences of surgery for gynaecological cancer

Abstract

It is important that a patient perspective is introduced to the identification and measurement of the outcomes of health care. The aim of this study was to use qualitative methods to examine the presence or absence of expectations prior to the experience of health care and the relationship between expectations, satisfaction and dissatisfaction in a group of women undergoing surgery in a large teaching hospital. Nineteen women with a diagnosis of gynaecological cancer were interviewed on two occasions, before and after surgery. A thematic analysis was undertaken. The results suggest that there is not a clear relationship between expectations and satisfaction. Women had different levels of expectations about different types of care and different aspects of care. Unfulfilled expectations did not lead to less satisfaction. The women were able to express satisfaction either with the care overall or with specific aspects of care, as well as being able to distinguish aspects of care with which they were dissatisfied.

Introduction

Health care should be evaluated in terms of what it achieves for the recipients of care. To a large extent, the frameworks within which objectives have been set, measures designed and evaluation implemented have been built on the knowledge and expertise of health care providers. Even patient satisfaction, the output measure most closely linked to the patient's perspective, is framed by the way in which providers have understood the structure, delivery and outcomes of health care. ¹ , 2, ^–3

There is growing evidence that incorporating the perspective of the recipients of care may lead to different emphases in terms of what is evaluated and how the evaluation might be undertaken. ^{4 , 5} To incorporate patients' perspectives, it is not necessary to change the assumption that if the objectives of a health care `intervention' are met, the outcome will be judged a success. What does need to be changed is the way in which objectives are identified and set so that patients' objectives – as well as providers' objectives – are incorporated into the evaluation framework.

One way of identifying the objectives of recipients of care is to ask them, a priori, about their expectations of the process and outcomes of care and then examine the links between stated expectations and subsequent judgements about the outcomes of care. In this paper, the concept of patient expectations and any link to outcomes is initially examined using the published literature. Then the results of a qualitative study of expectations of health care are reported. The results of this study are discussed in relation to the theoretical aspects highlighted in the first part of the paper. Finally, some conclusions about the way forward in health care evaluation are advanced.

Expectations and outcomes in the literature

It seems sensible to assume that what a patient or recipient of care expects to happen during health care and as a result of it (i.e. what they believe will or should happen) is likely to influence their opinion about the extent to which their expectations were fulfilled. Ross et al., ⁶ defined expectations of medical care as beliefs about things (needs, medical care, waiting time, etc.) and their attributes (urgent, curative, long, etc.). Whether the outcome is judged favourably depends on individual preferences and the certainty associated with the achievement of the outcome. The outcome of care from a patient's perspective has traditionally been measured as `patient satisfaction'. Thus the theoretical and empirical research reviewed below describes the relationship between patient expectations and patient satisfaction.

The Macquarie dictionary defines expectation as `a thing looked forward to; a prospect of future good or profit; the degree of probability of the occurrence of something'. Expectations are cognitive (i.e. they are a type of belief), are influenced by information and can be modified over time. Once established, beliefs manifested as expectations can affect attitudes.

The relationship between expectations and satisfaction has its strongest theoretical basis in psychology. Ross et al., ⁶ describe five theories which conceptualize the interaction between expectations and experience which result in a satisfied or dissatisfied consumer. Assimilation theory suggests that if expectations are not met, psychological discomfort will result. Consumers will attempt to reduce this discomfort by minimizing the differences between their expectations and experiences. Contrast theory proposes the opposite effect – that a consumer will magnify the differences. Assimilation‐contrast theory combines these theories and predicts that consumers will assimilate the differences between expectations and experiences when the disparity is low; the opposite will occur if the disparity is large.

Generalized negativity theory posits that disconfirmed expectations in either direction (either more negative or more positive) will result in a negative assessment of the service. Adaptation level theory proposes that an adaptation level is created by consumers based on their experience of the stimulus (or the situation). Once created, the adaptation level, which may be a level of expectations, will not change much except as a result of major influences. These theories have been tested in market research, but do not appear to have been the subject of any empirical research in health care.

A number of researchers have proposed value‐expectancy models ⁷ and alternative models such as discrepancy theory, fulfilment theory and equity theory ⁸ as possible explanations for the relationship between expectations and patient satisfaction. The theories described by Lawler ⁸ are implicit in a number of studies of patient satisfaction but they have not been tested extensively. All three posit that satisfaction is the difference between what is desired and what actually happens (although whether the desires are expectations or preferences or `what should be', is not clear. ⁹ Linder‐Pelz tested the value‐expectancy theory and the discrepancy and fulfilment theories. ⁷ Data, including patients' values and expectations, were collected from 125 first time patients at a primary care clinic. The overall conclusion was that very little satisfaction could be explained in terms of expectations and values despite there being some correlation. ⁷ Further, satisfaction was unrelated to fulfilment (defined as achievement of desired rewards) and inversely correlated with discrepancy (the better the outcome in relation to prior expectations, the more satisfaction expressed). A worse occurrence in relation to expectations did not lead to less satisfaction.

In a qualitative study of patients' expectations and satisfaction with health care for headache, Fitzpatrick and Hopkins ¹⁰ found that many patients waiting to be seen by a neurologist had difficulty articulating their expectations of the first consultation. They had their own ideas about the cause of their problem and some views about the type of examination and tests that might be appropriate, but their expectations regarding the outcome of care were limited. Fitzpatrick and Hopkins ¹⁰ also found that expectations were `fluid and emergent' and were modified in the light of experience. Patients did not necessarily feel that past encounters were reliable guides to this one and their expectations (hopes, fears, predictions about either process or outcome) were flexible. Expectations recorded before the consultation were tentative and as they did not vary significantly in strength, degree of concreteness or substance of expectations between subjects, were not useful in predicting responses to the consultation. The unformed nature of expectations may explain why McKay et al. ¹¹ in evaluating social work services found that although 80% of those whose expectations were fulfilled expressed satisfaction, 50% of those whose expectations were not fulfilled were also satisfied.

The results described above suggest that there is no simple relationship between what individuals expect before a consultation and any appraisal they might have post visit. There are no Australian studies in the published literature. Therefore, some exploratory research was undertaken to examine the presence or absence of expectations and the relationship between expectations and satisfaction in a group of women undergoing surgery in Sydney.

A qualitative study of expectations in health care

Method

This research was undertaken at Westmead Hospital, Sydney, with the permission of the ethics committee of the hospital. Women who attended the hospital for a pre‐surgical consultation between January 1997 and June 1997 were asked to participate. The surgeon, the registrar or the clinical nurse consultant explained the aims of the study to potential participants and permission was requested for the investigator (MH) to contact the participant and arrange an interview. Of 24 women who were approached, three were so ill they were admitted to hospital immediately after the consultation and two subsequently withdrew from the study; both women explained to the investigator that they felt too stressed to be interviewed. The study was described to the women as being conducted by a researcher from the University of Sydney who was interested in learning about their views on what they expected to happen in the hospital, their expectations about the impact of care on their health and the importance of various aspects of care to them. The 19 women with a diagnosis of gynaecological cancer who participated were interviewed on two occasions, before and after surgery for their cancer. Prior to the initial interview, the participant and interviewer (MH) together reviewed the information sheet and written consent obtained. Both interviews lasted between 40 and 70 min. In the week prior to surgery, 11 women were initially interviewed at home, eight at the hospital. The aim of the initial interview was to examine if patients had any `pre‐formed expectations' about either the process or outcomes of care. Using a semi‐structured interview format, women were asked about their expectations of the process and outcomes of care, any influences on the formation of their expectations and which (if any) of their expectations they saw as relatively more important. In addition, questions were asked about the decision making role the women saw themselves taking in the health care process. An interview schedule was developed which used the broad topic of expectations of health care to guide the interview. <sup>12</sup> The schedule contained no fixed wording or fixed ordering of questions, although the interview always began with a general question about the current experience of health care such as `Tell me how you came to be seen by Dr W (the surgeon)?'

The aim of the follow‐up interview, which occurred between 6 and 8 weeks post surgery, was to discover whether any changes to expectations occurred during the experience of health care, what these changes were and whether the women were aware of these changes. In addition, the interview explored the issue of what influenced the changes to these criteria. Finally, an attempt was made to clarify the `satisfaction' patients felt with the process and outcomes of care and whether they were satisfied in ways they expected to be.

All 19 participants had a diagnosis of cancer. The breakdown of their diagnoses was as follows: endometrial cancer, 11 (58%); ovarian cancer, six (32%); cervical cancer, one (5%); and vulval cancer, one (5%). Most women were aware of their diagnosis prior to consulting the surgeon (and hence being interviewed), although they were unsure of the extent of the cancer (i.e. whether lymph nodes or gastrointestinal tract were involved). The ages of the women ranged from 44 to 89 years. Five women were aged between 44 and 59 years and 14 women were aged between 60 and 89 years. Two women had previously been diagnosed with breast cancer (both in 1994).

Analysis

Tapes of the interviews were sent to a professional secretarial service to be transcribed. They were returned in both paper and computer disk format. The steps in the analysis of the interviews were as follows. ¹³

(1) Two people (the chief investigator and an independent person) listened to the tapes and read the transcriptions read to acquire a sense of the overall story being told by each participant. Both people then separately produced a brief description of the story being told or a list of issues raised during the interviews. These issues, as well as the framework provided by the interview schedule, were used to develop broad categories.

(2) The chief investigator re‐read the transcriptions and extracted `significant statements' from each interview. `Significant statements' can be distinguished from other parts of the transcription as being directly relevant to the category and/or opinions, preferences, assumptions, perceptions about events, staff actions or outcomes of the situation in question.

(3) The chief investigator attempted to interpret what was meant by each statement. This can also be described as moving from what was said to what was meant.

(4) Meanings were then organized into themes. Themes are recurring opinions, preferences, assumptions or perceptions. Themes were grouped together under the categories.

(5) Using these categories as a framework, the `results' were described in detail, using direct quotes from the interviews to illustrate the points being made. In this context, `results' mean the answers to the research questions outlined in the broad description of the interviews.

(6) From this `thick' description, the overall `story' was elucidated.

Results

The results are illustrated by direct quotes from the interview transcriptions. After each quote the participant is identified by number (from 001 to 019) and whether it was the first or second interview (1) or (2). The quotations from participants are drawn directly from transcription, any punctuation has been inserted by the transcriber.

First (pre‐operative) interview

Many women who participated in this study were either unsure about what to expect of the process and outcomes of care or unwilling to express their expectations:

No, I didn't know what to expect. Not at all. 004 (1)

I didn't have any expectations. It's all new to me. 008 (1)

Definite expectations tended to be about aspects of the process of care that women had some previous experience of (e.g. having a pre‐operative radiograph):

I expected to have some examinations. 001 (1)

I expect I will wake up and be very sore… I think you have to expect to wait [for the consultation]. 005 (1)

Other expectations were expressed in a tentative way, more as `hopes' than definite expectations. One of the reasons for the lack, or tentativeness, expectations may be because women did not want to be disappointed after previously expressing high expectations. In addition, many women felt that their limited experience with and/or knowledge about hospitals, medical procedures and other technical details rendered them less than competent to form expectations about such factors.

Such expectations as were expressed were about the technical (or tangible) aspects of care (such as being given adequate pain relief and being assisted in the first days after the operation):

I would expect to have quite a bit of discomfort and to be able to have a reasonable amount of relief from painkillers. 005 (1)

I expect to be hooked up to a drip and have some needles. 006 (1)

Women also expected that the results of the operation would be explained to them:

I am sure that [the doctor] will explain it [the results of the operation] to me afterwards. 009 (1)

They expected that help with self‐care would be available, but not be necessary for long:

I expect within two days to be up and moving around to the shower and everything on my own. 012 (1)

Very few expectations were expressed about the interpersonal (or intangible) aspects of care (such as communication or attitude of staff). Women may have judged that technical competence was the least that could be expected from staff, who were experts in their field, and having these expectations was not risking disappointment. However, they may have been unwilling to make the same judgements about interpersonal aspects of care without prior knowledge of the personnel involved.

Although they were unsure about what or how much to expect, some women had thought about the details of the care and in some cases discussed them with family or friends, often as a way of preparing themselves for hospitalization. For example, a number of women expressed the view that being provided with adequate pain relief should be part of modern health care. Most women anticipated that health professionals would give advice about when to return to `normal' activities. Finally, most women expected that doctors and other health care staff would provide practical information and advice if it was requested.

Both personal and experiential influences were obvious in the women's descriptions of how their expectations had developed. Personal characteristics were described as being important in the development of expectations about the overall outcome of the episode (i.e. the prognosis which they would walk away with after the treatment was complete), whereas experiences (their own or those of someone they were close to) were seen as influencing their expectations of the process of care, at least in the technical sense:

When I've had my other operations, I've always got up and going as quickly as I could… I've had other operations and it's been very uncomfortable for a while. 003 (1)

Everybody at work [had reassured me about the operation]…that's why I didn't worry because I thought they were all still walking around quite healthy…and my sister had cancer and she's all right… Yes, I haven't spoken to anybody who said they had a bad time. 010 (1)

However, women often hesitated to directly link such influences with their expectations of the coming experience. They felt that a diagnosis of cancer was an important limiting factor in the development of expectations about the process and outcome of care. Cancer was perceived as sufficiently different from other diagnoses that neither personal characteristics nor past experience were seen to provide good guidance to the development of specific expectations.

One issue where both personal and experiential influences were seen to play an important role was the expectations women had about the role they were likely to take in decision making about their health care. Some women expressed the view that personal characteristics were the most important influence on this role (e.g. if a woman was naturally quiet or naturally confident, they were likely to interact during health care decision making in a way that reflected these characteristics, no matter what their experiences were). Other women perceived that an important event or decision about their or a close relative's or friend's health care had made a profound impact on how they subsequently dealt with health care decision making.

Being cared for, having sufficient time to come to terms with what was happening to them, being treated in a timely manner, being given information and having some say in the decision making process were the aspects of health care that were emphasized as being important by the study participants. Pre‐operatively, women were able to describe by what criteria these aspects could be judged, but were unable to say what precisely they might expect from their coming health care. So, even where expectations were generally absent, unformed or not expressed, some features of care which were valued could be seen to be well‐formed and were able to be expressed. Some women described previous `good' and `bad' experiences of health care to illustrate their opinions about standards of care:

I've been to clinics before where they've left you exposed [describing her need for privacy]. 001 (1)

… I had that removed by Doctor [specialist]…and that was very pleasant. So discreet, it was wonderful. 008 (1)

A high standard of caring in previous health care experiences was perceived as occurring when the staff were polite, anticipated needs, were gentle and communicated well:

They took the fear away…because…you felt like you had known them for years …the way she [the social worker] put it [information] over to me, she didn't frighten me…and the way she did that was terrific [speaking about the pre‐admission clinic]. 001 (1)

Sufficient time was perceived to have been given when women had the opportunity to ask questions and offers were made of additional time for questions and/or discussion if necessary. Waiting to be treated was seen as a necessary evil. In the case of cancer, 1 to 2 weeks between the consultation and the operation was not perceived as being too long, although some women expressed frustration about the length of time they waited to see the surgeon.

There was considerable variation in the standards which women applied to information and decision making in health care. To some extent this may reflect differences in the amount or level of information they had received prior to surgery. Some women were more than happy with verbal information and many expressed satisfaction on being given additional written information which was available at a pre‐ admission clinic which some women attended:

…they explained everything to me and it was wonderful in the fact that I went home armed with all this knowledge of what was going to happen to me, what the likely outcomes were, the positive outcomes and the worst scenario I would find out at the end of this week…so I am well informed, very well informed, which would never have happened years ago. 012 (1)

Others felt the need for more information and indicated their willingness to look for it elsewhere (e.g. from their general practitioner or at the local library). Most women were discerning enough to recognize that these standards might change with circumstances. For example, many women described having more say in the decision making process when dealing with their GP or when having long standing health problems dealt with than under the present circumstances. Some women stated that they were always happy to let the doctor (or health professional) take the lead and make the decisions:

…I've been healthy all these…years… I've never needed any great decisions, not having to accept a decision like this… To me this is serious and I have to accept whatever advice I am given… I haven't got any say over it as such. 005 (1)

Others required more information, more questions answered, but were generally content to be convinced by professional opinion whilst a few perceived that they always investigated the problem thoroughly and in the end made their own decisions:

I would expect them [the doctor] to be the leader and then I would feel free to ask…they are the ones with expertise and the more advanced things. I think at times I have asked too much…once you have been through the mill and had one bad experience… I used to take everything for granted and never asked questions…but now I do. 001 (1)

Second (post‐operative) interview

As many women did not have or could not express expectations pre‐operatively, they had some difficulty deciding post‐operatively whether their expectations had been met. In addition, there were a small number of women whose post‐operative course was very different from the (relatively) smooth and gradual improvement in health experienced by most women. These women obviously did not experience what they had expected and their complicated recovery dominated their descriptions of their experiences.

Directly questioned, some women could describe what aspects of care were the same or different from what they expected, although they had not necessarily described these expectations in the pre‐operative interview. That is, their pre‐operative expectations were, at least to some extent, latent. Most of the expectations described as being fulfilled were about tangible aspects of care (e.g. length of stay, single room, help with self care) or outcome (e.g. return to normal activities, need for further treatment).

Aspects of care that were described as exceeding expectations included both technical (pain relief) and interpersonal ones (e.g. being taken notice, attitude of nurses, good care). Aspects of the experience that did not meet expectations (apart from the completely unanticipated complications mentioned above) were the under staffing of the ward, some technical problems (e.g. with drips and the epidural delivery of pain relief) and some aspects of the delivery and quality of the food. In addition, some women's expectations about the outcome (e.g. return to normal activity, the pathology results and the need for follow‐up treatment) were not met. Thus, women were more likely to have their expectations exceeded in aspects of care which they had not discussed in the initial interview. Possible explanations for this are that women set a low standard of care to avoid disappointment, or that they had expectations about what would not happen (e.g. that they would not be harmed, treated roughly or left in pain).

Most women described themselves as being satisfied with their recent hospitalization. The level of satisfaction could be seen to be related to the extent to which their treatment was `complete' at the time of the second interview. Those women who had required no follow‐up treatment or who had completed it were more likely to express themselves as completely satisfied with both the care and the outcome, whilst those who were still undergoing chemotherapy or radiotherapy were generally as satisfied with the `process' aspects of their hospitalization, but (understandably) unwilling to express overall satisfaction with any outcomes of care.

Interpersonal aspects of care contributed most to satisfaction whilst both interpersonal and technical (or `things' that happened to them) figured in the reasons given for dissatisfaction. In particular, satisfaction was influenced by care which was perceived as being individually tailored to their needs, when assistance was available and was given promptly, when needs were anticipated and when staff displayed attention to detail. Further, when staff displayed a caring attitude; when their manner was warm and they showed understanding of the woman's situation (including giving reassurance that things were going as planned and when family and friends were supported by staff), satisfaction was high. Satisfaction was also expressed when follow‐up care was given (e.g. community care organized, advice with post discharge problems was offered).

Many women expressed both general satisfaction and specific dissatisfaction. Despite the high overall level of satisfaction, all the women could clearly pinpoint aspects of the care with which they were dissatisfied. Dissatisfaction sometimes resulted from interpersonal failings (such as staff displaying attitudes of indifference or disinterest or when the attention from the doctors was not as good as women perceived that it should have been) but was more often described as resulting from technical failings (such as when a wound was not dressed, pain relief not sufficient, comorbidities such as diabetes and poor mobility not adequately dealt with, when women were not warned about the side‐effects of the epidural or other treatment or when treatment for an infection was delayed). In addition, the perception that the wait for pathology results was longer than it should have been, when miscommunication or organizational mix‐ups occurred (e.g. over discharge arrangements, advice to local doctors or a delay in the post‐discharge follow‐up consultation with the surgeon), when the `hotel' services were of a poor standard (i.e. the food was judged to be tasteless, not able to be eaten, in too large portions, the beds were uncomfortable and the surroundings were dull), also contributed to dissatisfaction:

I couldn't speak highly enough of the attention and everything that I got…except for the meals they served. They're enough to put anyone off. 005 (2)

There was one particular nurse who was very understanding and very nice…they looked after me well… I didn't like being in with four people, I would have liked two. 006 (2)

I have no adverse criticism concerning anyone or anything I had contact with. I was very impressed and I do not like hospitals… I thought the staff were incredible… I was overwhelmed at the high level of care… I wish somebody had warned me about that (headache associated with the epidural), not that there was anything I could have done about it. 014 (2)

I don't have anything against them. I think most of the nursing staff were really lovely… I got a terrible infection…and I wasn't very happy about that… I think there could have been a bit more care as far as things like doing dressings and things like that. That is the one area I was really dissatisfied with. 018 (2)

Women also expressed both satisfaction and dissatisfaction with the same aspect of care. This was expressed as overall satisfaction, but with particular `exceptions to the rule':

…the nursing care was very very good. I couldn't fault it…well really I was satisfied with everything…what had happened was that the evening nurse who had gone off must have touched the machine that was bringing me the epidural… I'm sure it was just a mistake. 007 (2)

An interesting aspect of the results was the extent to which individual women expressed specific expectations and the extent to which the fulfilment or otherwise of these expectations was related to satisfaction or dissatisfaction. Being in pain and having adequate pain relief were examples of expectations of an aspect of care which could be seen to be fulfilled and were likely to have contributed to the satisfaction subsequently expressed. This is how one women expressed herself:

I expect I will wake up and be very sore… I would expect to have quite a bit of discomfort and to be able to have a reasonable amount of relief from painkillers. 005 (1)

Oh the epidural was marvellous, and I certainly didn't have any pain, apart from when I coughed or sneezed unexpectedly when I wasn't prepared for it… 005 (2)

I couldn't speak highly enough of the attention and everything that I got. To me it was excellent… 005 (2)

However, the connection between expectations, their fulfilment (or otherwise) and dis/satisfaction was not always as obvious as the above example. Some women expressed expectations (e.g. about communication and ability to be self‐caring), only some of which could be seen to be fulfilled. Further, it was often the case that comments made by women which indicated overall satisfaction with care often did not relate directly to previously expressed expectations, whether they were subsequently fulfilled or not. One woman said:

I expect him to talk to me about it [the operation], to examine me… I expect within 2 days to be up and moving around to the shower and everything on my own. 012 (1)

After surgery, it was all over very quickly, and I was out of bed on the second day. 012 (2)

They were wonderful. When I needed them those first 5 days, they were there. I couldn't really fault the nursing at all. 012 (2)

Some tentative expectations, for example, about communication, were, when fulfilled, obviously related to subsequent satisfaction:

I expect to be told (the results) before anybody else I think. 010 (1)

Yes [they did explain the results]. And they assured me everything was good. 010 (2)

They spent the time and told me to ask for anything and I thought that was good… They had time to spend with you, time to talk… I wanted to ask them things and they never hesitated. 010 (2)

The fact that some expectations appeared to have been met did not always result in satisfaction with other aspects of care:

I expect to be told the results immediately. 018 (1)

Whilst I was in hospital (I found out the results), yes. 018 (2)

`Well, I think there could have been a bit more care as far as things like doing dressings and things like that. That is the one area I was really dissatisfied with' 018 (2)

Further, the fact that expectations were not expressed about a particular aspect of care did not preclude dissatisfaction with it.

Discussion and conclusions

Of the five theories proposed by Ross ⁶ only adaptation level theory may be supported to some extent by the findings of the current study which suggest that major influences (e.g. previous negative experiences) may change expectations. The results of this study do not support the theories developed by Lawler. ⁸

To a large extent, the results support the findings of Linder‐Pelz ⁷ that there is not a clear relationship between expectations and satisfaction. They also show that patients may not have expectations to begin with, thus mirroring the findings of the study undertaken by Fitzpatrick and Hopkins ¹⁰ which showed that patients' expectations were tentative and often modified by previous experience. In addition, some participants may have had difficulty in articulating or expressing their expectations, at least initially. Reflection on what was discussed during the first interview may have modified participants' responses to the second interview. Although the type and amount of information given to the women prior to surgery was not explicitly discussed in the interview, it was obvious that women varied in their knowledge of what to expect. This may also have influenced their responses.

In relation to the link between expectations and satisfaction, it was noted earlier that Linder‐Pelz ⁷ found that unfulfilled expectations did not lead to less satisfaction. This may be because individuals set a `minimum level' of expectations above which satisfaction is recorded. Another possible explanation is that unfulfilled expectations may result in dissatisfaction (as opposed to less satisfaction), which may require specific questions to elicit. This is borne out by the finding in this study that whilst most women were satisfied either with the care overall or with specific aspects of care, they were also able to distinguish aspects of care with which they were dissatisfied. Thus, satisfaction and dissatisfaction may need to be identified and measured separately.

This study was limited by its only being able to examine the experiences of women undergoing surgery for cancer. Men, people undergoing surgery for diseases other than cancer and those with chronic health problems may have a different perspective on the issues explored in this study. The participants were all being treated by the same surgeon and gynaecological oncology team. That the team were willing to cooperate with the research may indicate a high level of awareness of the issues raised in the research and a higher than average standard of care, thus biasing the results. Further, despite care being taken to distance the researcher from the providers of care, she may have been perceived as being closely allied to them. This may have led to a reluctance on the part of the participants to criticise those perceived to be in a position of power or who were judged to be doing their best under the circumstances. However, as they did feel able to criticize aspects of their care and appeared to be doing so in an honest manner, their comments can be used by the Gynaecological Oncology service to better understand the objectives of women and provide care which will meet these objectives.

Several areas for research have been highlighted by the results of this study. Discussing their expectations about health care may be one way in which patients express their objectives for care. However, objectives may be more fully elicited by directly enquiring about the reasons for seeking care and the goals or aims for both process and outcome of care. The results of this study also show that health outcomes (e.g. pain control, recovery from the operation and removal of the cancer) are not the only outcomes which mattered to women. Further research to identify, describe and, if possible, measure the range of outcomes of health care which count from the patient perspective is indicated.

Acknowledgements

I am especially grateful to those women who, in participating in this study, agreed to be interviewed at a difficult time in their lives. Acknowledgement is also due to the members of the Westmead Hospital Gynaecological Oncology Team who cooperated so enthusiastically with all aspects of the research. I would also like to thank Peter Sainsbury and Gavin Mooney who read earlier drafts of this paper and an anonymous referee for helpful comments. All remaining errors and omissions are mine.