Website of the Department of Health

The NHS patient series

The National Surveys of NHS Patients programme comprises a series of surveys designed to contribute to monitoring the performance of the NHS as seen from the patient's perspective. The Government committed itself to this programme in 1997 in The new NHS - modern dependable White Paper which proposed the introduction of regular surveys of patients and users to allow systematic comparisons of experiences over time and between different parts of the country.

The first survey in the series covered General Practice patients and was carried out in 1998. The second was a survey of hospital patients with Coronary Heart Disease (CHD), carried out at nearly 200 NHS Trusts in 1999.

The survey of cancer patients

This survey is the third in the series. Its overall purpose, as for the earlier survey of hospital patients, was to assess the quality of care as seen by hospital patients - in this instance with one of six types of cancer - breast, colorectal, lung, ovarian, prostate or non-Hodgkin's lymphoma. Over 65,000 patients drawn from over 170 NHS Trusts participated in the survey. Fieldwork began late in 2000 and was completed in the autumn of 2001.

Reporting the survey

The survey was designed on a scale sufficient to feed back results to each participating Trust, as well as to provide a national overview of patients' opinions of their treatment for each type of cancer.

Data for individual Trusts, grouped into Networks, will be available at the same time as this report. The data will include the results from patients surveyed in a particular Trust, in the context of national results and the results from the Network to which the Trust belongs.

This report presents the key findings from the survey for cancer patients in England from a national point of view.

Its focal point comprises the reported experiences and opinions of the six groups of cancer patients relating to the care and treatment they received from the NHS Trusts at which they were diagnosed with their illness, received their first treatment and attended most recently as an outpatient. Inevitably it concentrates on the differences between these groups in terms of their perceptions of the care received, their opinions of the doctors and nurses who treated them and in reported waiting times at various stages in the care pathway.

The care pathway does of course differ between patients with different tumours. Patients with different cancers present different symptoms; they require different investigations and different treatments are appropriate. For example it may be intrinsically more difficult to diagnose some cancers than others. Longer waits for diagnosis may not necessarily mean that the service quality is worse. Similarly some cancers and treatments may be intrinsically more painful than others. The readers of this report are likely to be only too aware that this is so and the report's findings should be interpreted in this light.

Comparison between Trusts

The report refers not only to differences between patients with different types of cancer but also to patients who received treatment, were diagnosed or attended as an out-patient at different Trusts. For comparisons between Trusts, the survey responses have been converted into 'problem' scores. Within the range of possible answers to a question, those which could be taken to indicate that the patient had a problem with the care or treatment that (s)he had received, were identified. Problem scores were then computed for each Trust for each of the questions where the answers could be interpreted in this way - approximately two thirds of the questions in the questionnaire. The Trusts were then ranked for each question from 'Best' - those with the lowest proportions of patients reporting a problem - to 'Worst' - those with the highest proportion of patients reporting a problem. This ranking was carried out on the basis of all patients as well as for patients with different tumours.

For any one question therefore, the 'Best' Trust for a particular cancer type could therefore be (and generally was) different from the 'Best' Trust for another group. Throughout this report there are references to the best performing and worst performing Trusts in particular respects. Some of the Trusts participating in this survey were excluded from these analyses because the number of responses received from their patients were too low to provide statistically reliable information. In particular, for three types of cancer - lung, ovarian and non-Hodgkin's lymphoma - the numbers of patients at individual Trusts were generally too low to allow the ranking from best to worst Trusts to be carried out. In instances where the best or worst performing Trust in the survey data was clearly an 'outlier' it has been excluded from the table in favour of the next-in-line.

The themes

Chapter 2 of this report presents the key findings of the survey in summary form. Most of the subsequent chapters of the report set out the survey findings in more detail under separate 'themes' - or aspects of the patient's relationship with the hospital, doctors, nurses and other NHS staff. The themes are similar to those used in the earlier national report on the survey of Coronary Heart Disease patients.

They are as follows:

Chapter 3 sets out the different profiles of the groups of cancer patients in terms of sex, age, ethnicity and social class.

Chapter 4 describes the extent to which patients moved between different NHS Trusts, for diagnosis, for first treatment and as out-patients.

Chapter 7 identifies the tests and treatments reported by patients with different tumours on the occasion of the first hospital treatment for their condition.

Chapter 15 draws together such indicators as there were in the survey data of differences in reported perceptions and experiences of demographic groups of cancer patients - men and women, different age and minority ethnic groups.

Chapter 16 gives a brief outline of the survey methodology.

The questionnaire was designed following a lengthy development stage, involving qualitative work among both patients suffering from different types of cancer and health professionals.

It was designed for cancer patients who might be at different stages of the care pathway. The aim, in so far as this was practicable, was to identify the stages and time lapses that might have occurred prior to the time at which the patients completed the questionnaire. It was also intended to capture patient experiences at all stages of the care pathway.

The questionnaire

There are six sections to the questionnaire:

A Questions relating to the hospital visit that led to the patient's being selected as a respondent; in particular whether this was the occasion of the first or a later visit.

B Questions relating to the first hospital treatment for the patient's condition, providing this had been within the past five years. The questions covered a number of 'themes' around which this report is structured - communication and understanding, patients' involvement in decisions, co ordination and continuity, physical comfort and pain management.

C Questions relating to the patients being discharged from hospital after the first treatment; the extent to which they considered they were given the appropriate advice and support after leaving hospital.

D Questions relating to the patient's referral to the hospital for diagnosis of their condition; the gap between referral and diagnosis and the way in which the diagnosis was given.

E Questions relating to the most recent outpatient appointments; whether these were subject to postponement, whether they were frequent enough and the way in which the patient felt (s)he had been treated.

F General demographic (age, sex, etc) and related questions.

The tables

Of the six groups of cancer patients included in the survey, breast cancer patients, by a clear margin, formed the largest group. They constituted 39% of the total sample. It also proved to be that across some, but not all, aspects of the findings, the profile of opinions and reported experiences of this group of patients was quite distinct from the equivalent profiles of patients with other types of cancer. The appropriate comparison, therefore, for these findings is between breast cancer patients and other patients, not with all patients, of whom they form such a large subset. The tables showing analyses by tumour type are generally presented in the report in this format. It was felt preferable to retain this format for these tables consistently throughout the substantive chapters of this report even where there were few differences between breast cancer and other cancer patients.

The numbers of patients with each type of cancer are set out in Chapter 3, Table 3.1. These are the bases for percentages for the tables showing analysis by tumour type, after allowing for small numbers of missing cases. In the interests of clarity of presentation they are not repeated at the foot of each of the tables. Where the tables are necessarily based on a sub-sample - eg those patients visiting their GP before the first hospital appointment, or those patients with an out-patient appointment in the last two years - then the bases for percentages are shown at the foot of the tables.

Where differences between groups of cancer patients are commented on in the text, these are generally statistically significant at a level of 95% probability.

The symbol * in a table signifies less than 0.5%.

Apart from analyses by tumour type, there are a number of tables in the report which indicate the 'problem' scores at the best performing and worst performing Trusts (see the relevant paragraph in this Introduction). These analyses are derived from a database restricted to 164 out of the 172 Trusts that participated in the survey. (Response at the eight omitted Trusts was too low to provide statistically reliable information.) These tables show the best and worst Trusts for three types of cancer patients only - those with breast, colorectal or prostate cancer. (There were insufficient numbers of patients of the other three cancer types to allow this form of analysis to be carried out.) Even within the nominated three cancer types there were some Trusts with fewer than 50 patients with a particular type of cancer, for a particular question. In these cases the Trust would be excluded from the best:worst ranking.

Fieldwork therefore did not begin until after the publication of the National Cancer Plan in September 2000. However, the period of reference for the survey on which patients participating in the survey were asked to focus was, for the great majority of questions, that prior to July 2000. The Cancer Plan includes targets for reducing waiting times, improved standards for cancer services and increased staff numbers and training. This report therefore offers a benchmark from which the implementation of the National Cancer Plan can be monitored.

Following the publication of the National Cancer Plan, NHS Trusts were grouped into cancer 'Networks' to improve the co ordination of cancer care. The Trusts participating in this survey belong to 34 different Networks. Typically a Network would include four to six Trusts in relatively close proximity.