THE STATE OF READINESS OF READ CODES VERSION 3 FOR WIDESPREAD USE IN THE NHS

43. In 1992, the NHS Executive began a programme of work in collaboration with the clinical professions to expand the Read Codes to cover the clinical terms used by health care professionals in the hospital and community health sector. The work comprised two main phases:

• the clinical terms projects which began in 1992 with the aim of developing a computerised thesaurus of Read coded clinical terms and groupings; and

• a subsequent programme of work to refine, further develop and test the Read coded clinical terms which began in 1994.[55]

44. In May 1995, the Centre began a programme of work to implement Read Codes Version 3 in the NHS. This programme included:

• a partnership with 12 NHS trusts to pilot Read Codes Version 3 in live use;

• in­house tests to assess strengths and weaknesses in the codes;

• external work to assist with the development of the codes using data from an acute hospital; and

• a retrospective review of the queries reported by Pembrokeshire and Derwen NHS Trust during the first year of testing Read Codes Version 3 in use in a clinical environment.[56]

45. In January 1998, there were 12 NHS sites in the hospital and community health sector known to the NHS Centre for Coding and Classification to be testing and using Read Codes Version 3 in clinical information systems.[57]

46. The Committee asked the Chief Executive about the viability of Read Codes Version 3 in the light of this low level of usage, of concerns about the work needed to complete Version 3, and the view of the Royal College of Nursing that they did not recommend Version 3. We also asked the Executive whether Version 3 was now a complete service strength product capable of doing its job in a NHS acute unit.

47. The Executive told us that they had made real progress in developing Version 3. They had made further progress in integrated clinical systems in different hospital settings. They had reached a very detailed agreement with the medical profession about how they were going to handle security and confidentiality in those systems. And they had been managing the roll out of the unique patient identifier that gives every member of the population a unique number for use in health records. These were the essential building blocks of good clinical systems.[58]

48. The Executive told us that in the right circumstances, with the right clinical systems, Version 3 was "pretty well service strength", and had been so for some six to nine months.[59] They added, though, that Version 3 would require continuous development, and that it would only function effectively with the appropriate clinical systems in place.[60]

49. The Executive said that they had not seen the views of the Royal College of Nursing, but they pointed out the Medical Information Group had spoken highly of the product, and people in the NHS who were using it could see its tremendous potential.[61]

50. We asked when Version 3 would be in operation. The Executive told us that, provided the clinical systems were in place, Version 3 should be implemented over the next five years. When we asked them why it would take a further five years to implement, the Executive explained that implementation was linked to wider developments in information technology. The NHS had to:

• Modernise its information systems;

• Invest in clinical information systems that would support the capture and use of clinical information in hospitals and general practice;

• Use technology that would allow communication electronically between hospitals and general practitioners; and

• Use a standard clinical vocabulary such as Read Codes to support all of that. [62]

51. The Executive had not yet set any targets for investment in suitable host clinical information or for the take­up of Read Codes Version 3 in the NHS. Such decisions were taken locally. At the time of the National Audit Office examination, they intended to reconsider the question of mandating Version 3 of the Codes for use in the NHS in the context of the new Information Management and Technology strategy.[63] We therefore asked whether the Executive intended to mandate Version 3 of the Codes to the NHS. They told us that this would be a decision for ministers. Ministers wanted effective communication between general practitioners and hospitals, and the modernisation of the relevant technology between the two. To do that, they would need a standard vocabulary such as Read.[64]

52. There has been no overall independent assessment of the value and effectiveness of the Read Codes.[65] The C&AG considered it essential to the future development of a clinical coding system that the NHS Executive commission an independent evaluation, or evaluations, to confirm that the Read Codes would meet the Executive's objectives within their Information Management and Technology strategy. In his view, the Executive would need to confirm that the investment represented value for money and that the Read Codes:

• were usable in a live environment;

• could support the production of statutory (central) returns; and

• could support the clinical information needs of health care professionals and the business information needs of managers.[66]

53. We therefore asked whether, given the conflict of professional opinion about the very validity of the Read Code system, the Executive accepted the need for an assessment of the whole system by an independent panel of experts. The NHS Executive outlined a number of benefits from the use of systems such as the Read Codes, including:

• cash-releasing efficiency improvements, such as savings in clerical time;

• non-cash releasing efficiency improvements such as improved accuracy; and

• improvements in patient care, such as better analysis of referrals which made it possible to tailor services.[67]

However, they confirmed that they planned to commission independent review.[68]

54. We asked who the Executive envisaged undertaking the review, and what the terms of reference would be. They told us that they would be tendering for the work on a European basis and that they expected the review to be complete by December 1998. They undertook to keep the Committee informed at each stage on how the review was progressing, and in particular to let the Committee know who would undertake the work, the terms of reference and the outcome.[69]

55. We asked whether the review would consider alternatives to the Read Codes. The Executive told us that that there were very few alternatives, but that it would be it very surprising if ministers did not want to look at them. They had already had some discussions with ministers; and the outcome of those discussions would become known when the Information Management and Technology strategy was published.[70]

56. We asked whether the Executive should not wait for the outcome of the review before doing anything more or spending any more money on the Read Codes. They told us that, if they were to maintain their leading edge in developing clinical language, they could not allow a period of planning blight while the review was being carried out.[71] We asked how much more money the tax payer would need to put into the system, against the possibility that it might not work. The Executive told us that Ministers would need to agree what the total expenditure would be on updating clinical systems needed to run Read Codes.[72]

Conclusions

57. Eight years after the Read Codes were purchased and three years after the clinical terms projects ended, Version 3 of the Read Codes is being tested and used in only 12 NHS hospital sites. If the benefits of Read Codes are as good as the Executive suggest, progress has been very disappointing; and if Read Codes are to be implemented nationally, it is likely to be at least a further five years before the Codes are implemented throughout the Service.

58. We are concerned about the conflicting views of key groups in the NHS, such as the NHS Executive, the Royal College of Nursing and the Medical Information Group, over whether Read Codes should be used. We welcome therefore the NHS Executive's commitment to an independent evaluation of the Read Codes, as recommended in the C&AG's report.

59. The review needs to derive its authority from its scope, the rigour of its methodology and the expertise and transparent lack of conflict of interest of those conducting it. We expect the review to:

• take full account of assessment, testing work and professional opinion to date;

• assess benefits, costs and value for money;

• produce a realistic assessment of the likely timescales for development, implementation and usefulness of the systems, taking into account the linkages with investment in NHS information systems and the new NHS Information Technology Strategy;

• assess the risks and develop strategies for managing them.

It is important that the NHS Executive should not feel locked into using the Read Codes because they have already spent so much money on them. The review should be prepared to look hard at alternatives.

60. Clear lessons have emerged from the development of the Read Codes, about setting out business cases about project and programme management and about implementation of systems across the NHS. We look to the Executive to apply these lessons to other areas of their Information Management and Technology Strategy.

56   ibid, paras 2.32-2.34 and Figure 6 Back

57   C&AG's report, (HC 607 of Session 1997-98), para 2.34 Back

58   Q125 Back

59   Qs 55-60 Back

60   Q78 Back

61   Q27 Back

62   Qs 145-147 Back

63   C&AG's report, (HC 607 of Session 1997-98), para 2.35 Back

64   Q33 Back

65   C&AG's report, (HC 607 of Session 1997-98) para 2.38 Back

66   ibid, para 35 Back

67   Evidence, Appendix 1, pp 23-25 Back

68   Qs 34-35, 91 Back

69   Qs 34 and 35 Back

70   Qs 200-204 Back

71   Qs 91-95 Back

72   Q139 Back