Electronic patient record

Electronic patient record

The advancement of computer systems in the USA has demonstrated the potential advantages of the introduction of EPRs, with many of the benefits related to the introduction of electronic prescribing systems within them. Whilst introduction of these systems into the UK is still very limited, the previous government’s stated ambition was to have EPR systems, including electronic prescribing systems, in place in all acute hospitals. The report from the Treasury, the Wanless report, underlined the importance of investment in IT, which could be taken as supporting the development of the national programme for IT. It identified that:

one of the major benefits to accrue from such systems arises from physician order entry which requires the doctor to input requests for new medicines, X-rays or other investigations into the system directly. This physician order entry has been a feature of most UK applications whereas the health system in the USA, delivered as it is by independent private physicians, does not lend itself to mandatory requirements for physician order entry. As a consequence, despite having advanced computer systems, many of the advantages of an electronic record are missed due to lack of physician involvement with the system.

The use of EPR systems has shown a number of benefits in terms of improving communication and reducing risks to patients. A patient safety internet site estimated that computerised physician order entry could lead to the avoidance of 522 000 serious medication errors each year in the USA. This has been demonstrated in terms of improved legibility and completeness of prescrip-tions and dramatic improvements in patient care, with reduction in risks associated with inappropriate dosing or drug choice. Advanced systems that include clinical decision support with dose and interaction checking or checks of the appropriateness of the prescription allow review of the prescrip-tion at the point of decision-making.

The sentinel paper that highlighted the potential of e-prescribing to reduce errors was published by Bates et al. in 1998.A more recent systematic review reports a relative risk reduction in medication error rates in 23 out of 25 systems of between 13% and 99%. Six of the nine studies reviewed looking at the effects on potential adverse drug events showed a relative risk reduction of between 35% and 98%.

Thus there are a number of different ways in which e-prescribing systems can reduce the incidence of medication-related error. However, caution is required: many of the studies published that describe the benefits can be criticised for poor methodology and outcome measures. Also, it should not be forgotten that there are also ways in which systems can introduce new errors or increase error rates. These must be taken into consideration during design and implementation.

The main areas of benefit for a well designed e-prescribing system include:

·      Legibility and completeness – prescriptions are legible and complete. Structured pathways for ordering medicines ensure that full descriptions are used, doses use appropriate units of measure and frequencies are selected for individual medicines. For example, the word ‘unit’ can be forced for insulin doses, micrograms can be forced in place of mg, addressing known errors with hand-written scripts.

·      Decision support – this can be utilised in a number of different ways:

·      Passive decision support: systems can be configured to ensure that prescribers are guided to select the most appropriate prescription content, namely route, unit of measure, frequency and other supporting requirements, such as appropriate rates of administration or diluents. Routes of administration can be limited; for example, the intrathecal route can be excluded for selection for all vinca alkaloids.

Much of the reduction in error rates reported by Brigham and Women’s Hospital has been due to this type of support.

·      Active decision support: active checking of prescriptions can be undertaken using prebuilt third-party or local rules. When a rule is ‘broken’ an active alert will fire a warning to the prescriber. Areas that can be supported include allergy checking, drug–drug interaction checking, dose range checking, therapeutic duplication and local rules to support, for example, meticillin-resistant Staphylococcus aureus prophylaxis and antibiotic course length review.

·      Formulary management – this can be used in a number of ways. It can be used to ensure that medicines prescribed are available, thus potentially reducing delays in administration. It may also be used to limit access to specific medicines on safety or clinical grounds, supporting compliance with local policy. For example, improvements in the use of antibiotics have been reported, including a reduction in misselection and excessive doses. It may also be used to support improvements in prescribing appropriate prophylaxis such as for heparins⁵

·      Communication – prescriptions are available where and when they are required as they can be accessed via many terminals or locations at the same time. Current paper processes require prescriptions to be moved around, meaning that they may not be in the correct place or are lost. E-prescribing promises much in terms of improving communication between different care sectors in the future and supporting medicines reconciliation.

·      Administration: the introduction of bar code checking to support the correct identification and recording of medicines administered has been reported to have reduced the incidence of error by up to 80%.

·      Introduction of new errors: there have been a series of reports outlining problems with e-prescribing systems. Many of these have been due to poor implementation or configuration decisions, often where clinical input has been weak or planning inadequate. A case in point is the implementation of a paediatric system in the USA that initially increased mortality rates. Other problems vary but include: (1) poor system design, particularly relating to decision support; (2) the process flow used within systems; (3) poor screen design; and (4) a lack of decision support. In particular, if alerts are not relevant or fire off late in the prescribing process, it is more likely that they will be ignored and important warnings missed. There is a fine balance to be achieved to ensure that overalerting is avoided.

Poor screen design may repeat many of the problems seen with poor hand-writing, for example, the use of abbreviations, truncated descriptions and inappropriately wrapped text. Many basic system errors that can be mitigated via system design or configuration are outlined in the hazard frameworks published by the e-prescribing programme at NHS CFH and in user interface guidance for medicines which can be found at the common user interface program.^{31, 32} The NPSA has also published guidance on the safe on-screen display of medication information which illustrates some of the basic design elements that can be used to reduce risk in this area.

The Leapfrog Group in the USA has continued to champion the introduc-tion of e-prescribing (or computerised physician order entry (CPOE), as it is known in the USA) into hospitals. Over the past 10 years it has developed a tool to review individual system implementation of CPOE at hospitals to see whether key patient safety benefits are met.¹⁴ This is being used by various accreditation bodies in the USA to support patient safety benefits at hospital level and has been funded by the Agency for Healthcare Research and Quality. At the time of writing, summary results of the initial use of this tool have yet to be published. Personal communication with the author has suggested that implementation of CPOE (and thus the benefits) can be very variable between sites using the same system; this highlights that local configuration decisions and ongoing development are key to supporting system benefits realisation. Work is under way to see whether the tool might be used in the UK to support ongoing system review and development at trust level. Initial results have been promising and have demonstrated that Anglicisation is entirely possible.

The benefits of e-prescribing are increasingly being sought by UK hospitals to support the quality and safety agenda of which improvement in the use of, and the management of, medicines forms a large part. Communication between different care providers by facilitating medicines reconciliation and transfer of information are areas that are being highlighted as being particu-larly beneficial to patient safety. The Care Quality Commission report iden-tifying the challenges with managing patients’ medicines after discharge from hospital highlights that IT could be used to facilitate improved information transfer between care settings.

A significant number of medication-related errors due to poor hand-writ-ing and incomplete information are likely to be reduced with the introduction of e-prescribing. The 2009 General Medical Council report highlighting the number and type of errors in prescribing adds further weight to organisations’ need to introduce systems to facilitate improvements in prescribing practice.

E-prescribing is one of the most, if not the most, complex areas of health IT to develop and implement – both technically and culturally. One of the significant challenges for the introduction of IT is the cultural change required for the successful implementation of these systems. There are a number of difficulties in getting buy-in from a range of different professional groups. Successful implementation requires strong clinical champions, a commitment from the highest level in the organisation and a clear expression of the clinical benefits that patients will gain from its introduction.

Prescribing is an activity that affects almost every patient admitted to hospital and requires the active participation of the majority of hospital staff. It is also an area that must be failsafe and there is no leeway for mistakes to be made. It is an area that, to many, seems too complicated to implement and often most enthusiasm for implementation comes from pharmacy. Given these challenges with the introduction of e-prescribing, NHS CFH commis-sioned a report to identify the lessons learnt from successful and unsuccessful UK implementation of e-prescribing. This highlights the main lessons and areas that must be addressed to facilitate clinical buy-in and successful system rollout. The publication also contains six briefing documents aimed at the various groups and clinical professions that must be involved, demonstrating the size of the engagement challenge.

It should not be forgotten that, although e-prescribing is generally one, if not the main, focus for delivery within an EPR, support for clinical pharmacy and medicines management services should also be incorporated. Work has started to define the core standard requirements laying the foundation for a pharmaceutical care record that can be shared across and beyond the phar-macy profession within an overall individual patient record. The profession must develop its requirements to support developments and advances in practice called for in the 2008 pharmacy White Paper.³⁸ It needs to ensure that information is included within the overall patient record and input is visible to the wider healthcare team. Whilst EPRs offer many opportunities to develop pharmacy practice by freeing time to focus on clinical input, they may also represent a threat if the automated elements are seen as staff-saving.³⁹ Systems are not a panacea for all medication-related problems and this is not always understood. Failure to understand the opportunities and threats from the introduction of systems may create challenges in terms of staffing and service development, of which the profession must be aware if a reversal in clinical input is to be avoided.