Patient Safety - Health Committee Contents


7  Technology and patient safety

155. All forms of healthcare involve the extensive use of technology; and medicine continues to change dramatically as a result of both the refinement of existing types of technology and the development of new forms. Innovative clinical techniques (in diagnosis, medication and surgery alike) have transformed many aspects of clinical practice, reducing the risks and increasing the benefits of almost all types of medical intervention.

156. However, the complexity and power of modern, high-technology healthcare brings with it increased scope for things to go wrong, and for adverse events to have very harmful consequences. Moreover, the introduction of new clinical techniques can pose a risk to patients when their introduction is not accompanied by appropriate training or they are not properly integrated into clinical pathways. Professor Bryony Dean-Franklin, Director of the Centre for Medication Safety and Service Quality at Imperial College Healthcare NHS Trust, outlined in her evidence some of the main potential drawbacks associated with technology.

  • Over-reliance on the technology to check for errors, and assuming that the technology is more effective than it actually is, and so decreasing personal vigilance
  • Development by staff of work-arounds to avoid safety features
  • New types of error introduced, eg errors when selecting from pull-down menus
  • Introduction of additional steps into work processes
  • Failure of the computer system
  • Deskilling [of healthcare professionals]
  • Assumption that the technology is more advanced than it is—eg assuming that a computer system includes decision support where none exists.[169]

157. Nevertheless, properly used, technology can greatly improve patient safety. Some new technologies are less invasive and reduce the potential for harming patients. Other forms of technology might reduce the types of error classified as "slips and lapses" or "failure to carry out a planned action as intended". The automation of various processes in healthcare could reduce, or perhaps even eliminate, a number of errors that are attributable to causes such as failures in memory and lapses of attention on the part of both clinical and non-clinical healthcare workers. This could apply to areas where the accurate carrying out of tasks such as checking and measuring is crucial to patient safety.

158. A number of such innovations have been successfully piloted in parts of the NHS, but witnesses were concerned that their adoption throughout the service was unnecessarily delayed.

Automated decision-support

159. We heard from Dr Olga Kostopoulou, of the Medical Decision Making Research Group at Birmingham University, that a possible solution to the problem of diagnostic error in general practice was greater use of automated decision-support systems, which can help GPs ensure that they do not miss less common diagnoses.[170] We also heard that automated prescribing-support systems, such as drug interaction programmes, offer great potential for making prescribing safer and are already being successfully used by many GPs. As we saw during our visit to Charing Cross Hospital, such systems also have an application in the acute setting, although we were concerned to find in this particular case that the alerts had been switched off.

160. Automated decision-support systems would be even more useful if National Institute for Health and Clinical Excellence (NICE) guidance could be integrated into them. The Chairman of NICE, Professor Sir Michael Rawlins, gave us an encouraging account of the development of NHS Evidence for this purpose—although the system does not yet allow for the automatic interaction of NICE guidance with patients' records.[171]

Automatic Identification and Data Capture

161. Automatic Identification and Data Capture (AIDC), including barcoding, has the potential to reduce errors during routine tasks requiring accuracy, such as checking and measuring.[172] The benefits of AIDC technology were outlined in 2007 in the DH's Coding for Success: Simple technology for safer patient care (see Box 13). During the inquiry we looked at a number of areas where barcoding could bring such benefits. Box 13: Potential benefits of Automatic Identification and Data Capture technology, such as barcoding
Verification - a major application of AIDC is to verify the identity of an item, person or procedure and link this with the member of staff involved in patient care. A hand-held computer with a built-in scanner can be programmed with protocols for procedures such as blood transfusion […]

Data capture - there are many situations where serial numbers or reference numbers need to be entered into electronic records. Using AIDC to enter the information eliminates the risk of manual keystroke errors […]

Supply chain issues - effective track and trace of goods improves stock control so the right supplies are available in the right place at the right time. Unique product codes can be used on individual high-value items as an anti-counterfeit measure. Linking patients to the supply chain—by recording product information on patient records—can help to identify batches where a patient or patients have had an adverse reaction.[173]

162. A number of schemes in New Zealand have shown the benefits of AIDC. The New Zealand Medication Safety Project has produced benefits by introducing bedside verification of prescriptions using barcodes.

163. In Auckland we met Professor Alan Merry, who has developed the use of barcoding technology for double-checking anaesthetic drugs in operating theatres with considerable success. This is now being piloted in the UK, as the DH told us:

    The NPSA, in collaboration with the Royal College of Anaesthetists is currently undertaking a qualitative study into the feasibility of introducing double checking procedures to improve the safety of the administration of injectable anaesthetic drugs. As part of this study bar coding technology that has been developed by Professor Alan Merry in New Zealand is being piloted in two hospitals in the UK.[174]

164. During our visit to Charing Cross Hospital, we saw how barcodes could be used to automate drug dispensing and administration. The hospital uses a pharmacy-based dispensing robot, which is essentially an automated picking mechanism in a closed cupboard, handling medicines by means of barcodes. Use of the dispensing robot avoids picking errors (e.g. picking the wrong drug because of identical packaging or similar drug names)[175] and allows for better use of space (as the robot requires less room to move around than a person does). While there is still some scope for errors to occur and it remains essential that staff continue to check and do not assume that the robot makes any error impossible, use of the robot has been proven to reduce the level of errors overall.

165. The hospital also uses CII Safe, an automated dispensing system utilising barcodes for the administration, documenting, storage and ordering of controlled drugs, such as morphine.

166. We also saw at Charing Cross Hospital a pilot project for electronic prescribing, automated dispensing on the ward and barcode verification of patient identity (the ServeRx closed-loop system). There is still scope for error to occur; for example, in the USA problems have occurred where these systems have been used by an unskilled workforce, accustomed to assuming that the system is always right. Nevertheless, by almost completely doing away with handwritten prescriptions, the Charing Cross Hospital system helps eliminate illegible prescriptions, lack of information from the prescriber, and nonsense prescriptions—dramatically reducing drug administration errors.

167. Professor Mike Murphy heads a team at the Oxford Radcliffe Hospitals NHS Trust that has used barcoding to transform how blood for transfusion is handled. He explained that this project had been a proven success, in terms of both safety and saving money—yet it still had not been implemented throughout the NHS. His team had been funded by Connecting for Health to write a national specification, but this was still being trialled (at the Mayday Hospital in Croydon). Professor Murphy told us that "we have all been incredibly frustrated about how slow [the pilot] has been".[176] He identified underlying causes of delay as follows:

    [T]here are huge financial barriers […] Finance for the equipment is one; finance for the training is another. There are IT connectivity problems within hospitals, between patient administration systems and laboratory systems and other systems within hospitals. We would like to use the NHS number [for patient identification purposes], but we have problems with that […][177]

168. We questioned the DH about the failure to make more progress. The Department has recommended that the GS1 system[178] should be adopted as the universal standard for AIDC technology throughout the NHS in England, both for manufactured products and for coding systems used within healthcare settings, such as patient identification codes on wristbands.

169. The DH had promised to review progress in the implementation of Coding for Success by the end of 2008. However, we were told by Ann Keen MP, Parliamentary Under Secretary of State at the DH, that the NHS Procurement e-Enablement Programme, which had not been envisaged when Coding for Success was published, was "starting to drive uptake of coding" and so the Department had "decided that it was no longer appropriate to undertake a review in 2008".[179]

Electronic Patient Record

170. As we noted in a previous report:

    Storing and transferring patient information electronically has the potential to significantly reduce clinical errors and improve patient safety as well as allowing clinicians to communicate more quickly and accurately and to identify relevant information more easily.[180]

171. During that inquiry we saw how the successful implementation of the Millennium Patient Administration System at Homerton Hospital in London had made services safer for patients. More recently, when we visited St Thomas' Hospital in London, we saw the application of electronic records in the A&E care setting, where the Symphony system (specifically designed for emergency care) is in use. In both these cases, however, their systems are free-standing and not yet networked with the rest of the NHS, not even with local general practices—at St Thomas' we heard that the only way for patients' GPs to see the records is for them to be printed off in "hard copy" and put in the post. Networking of records between all care settings is vital if computerisation is to facilitate better continuity of care between acute and primary care sectors to address acknowledged risks such as poor medicines reconciliation.

172. The creation of an Electronic Patient Record (comprising a Summary Care Record and a Detailed Care Record) that is universally accessible throughout the NHS has been the objective of the Care Records Service element of the NHS National Programme for IT. We noted in our report on the Electronic Patient Record that implementation of the Care Records Service had fallen significantly behind schedule, for a number of reasons. And the NAO has since reported that:

    Current indications are that it is likely to take some four years more than planned—until 2014-15—before every NHS Trust has fully deployed the care records systems.[181]

Technical solutions to preventing catastrophic disasters

173. We heard from Professor Brian Toft about his inquiry (on behalf of the CMO) into the death of Wayne Jowett, a patient who died in 2001 when an anti-cancer drug (Vincristine) was mistakenly injected into his spine instead of into a vein. In summarising his findings, Professor Toft explained that there was a simple solution to this problem:

    You can connect any Luer[182] syringe to a spinal needle, which meant that Vincristine could be put into his back. Once they got to that stage, that was it, finished. One of my recommendations was they should change the technology and design a spinal needle to which a Luer syringe could not be connected. That seemed to me to be a reasonable way to progress. Since that point in 2001, and we are now in 2009, we still have not got that connector. There has been a lot of reasons for this. In the first instance, although I have had no close supervision of it, there is the commercial interest—who will be the manufacturer who would be prepared to put money into creating such a syringe—and, then, will the medical profession buy it? The next thing, of course, would be the designing of the syringe—how would we design it—and then a whole range of other issues surrounding whether professionals will be prepared to use it when they are used to using their own methodologies. Would the needle work in the way it is supposed to do? What about the law of unintended consequences? In systemic terms, will the creation of this create another problem in another area and lead to a patient death for another reason? And so forth. It has been those kinds of things. I did ring the Department of Health yesterday and ask them how we were doing, as the saying goes, and they said they had some prototypes that they were working on now and they hoped, within the not too distant future, they would have a suitable solution to that problem. Of course, I cannot say how long the future is.[183]

174. The NPSA has been consulting on a draft "purchasing for safety initiative", requiring healthcare organisations to use neuraxial (spinal) syringes / needles and infusion systems with connectors that will not connect with intravenous devices. It is intended that NHS bodies will be required to act on this by 2011. The intention is apparently to stimulate both the medical devices and pharmaceutical industries to produce this type of connector, possibly ahead of the new International Organization for Standardization Industry Standard (ISO) for small-bore (non-Luer) connectors for neuraxial and other medical devices.

175. Unfortunately, no manufacturer has to date begun production of such a simple yet life-saving gadget.[184] This may be in part because the ISO standard has still yet to appear nearly a decade after a European Standards Agency Task Force recommended in 2000 that Luer connectors be restricted to intravenous and hypodermic devices.

Conclusion

176. While the potency and complexity of modern technology mean that it carries great potential for harm, it can also make a major contribution to patient safety. During the inquiry we took evidence about a number of technologies which could make significant improvements to care but which were being implemented far too slowly.

177. Automated decision-support systems can help improve patient safety, notably in primary care. We note the slow progress made in integrating National Institute for Health and Clinical Excellence guidance into such systems and recommend that a timetable be set for achieving this.

178. Electronic prescribing-support systems should be introduced throughout the NHS and set up with the alerts feature appropriately configured.

179. Automatic Identification and Data Capture technology, such as barcoding, has the potential to reduce significantly certain types of error. Impressive pioneering advances, such as those in relation to blood transfusion at Oxford Radcliffe Hospitals NHS Trust and to medication at the Charing Cross Hospital, have been made in this respect, but we have grave concerns about their slow uptake across the NHS. We are concerned at the DH's decision not to review progress on Coding for Success. Its reasons for not doing so are unacceptable in view of the slow progress to date.

180. The continued delay in the Electronic Patient Record also represents a huge missed opportunity to improve patient safety by improving the communication of clinical data (particularly between care settings), which would reduce administrative errors and facilitate better continuity of care.

181. We are alarmed at the lengthy delay in implementing Professor Toft's 2001 recommendation regarding the development of spinal needles that cannot be connected to a Luer syringe. It is totally unacceptable that an identified and simple technical solution to a catastrophic problem should take so long to be put into practical use. The Chief Executive of the NHS must explain why this delay has taken place and ensure that such delays never occur again. It is unacceptable that the NHS does not have a mechanism to ensure that changes such as this, which impact seriously on patient safety, occur in a timely fashion.


169   PS 86 Back

170   Q 647 Back

171   Qq 857-858 Back

172   AIDC technology encompasses the use of barcodes, radio frequency identification and other machine-readable codes to identify quickly and accurately an item or process. Back

173   Department of Health, Coding for Success: Simple technology for safer patient care, February 2007, para. 2.4, p 7 Back

174   PS 01A. See also Ev 269. Back

175   We heard from Professor Brian Toft about look-alike medicines (where different drugs or dosages are similarly packaged) and sound-alike medicines (drugs with similar sounding names) and the apparent reluctance of manufacturers to change their packaging (Qq 221, 266). Back

176   Q 302 Back

177   Q 308 Back

178   "GS1 is a neutral, not-for-profit standards organisation developing global supply chain standards and facilitating the adoption and implementation of such standards" (Ev 181 [GS1]). Back

179   PS 01B Back

180   Health Committee, Sixth Report of Session 2006-07, The Electronic Patient Record, HC 422-I, para 1 Back

181   National Audit Office, The National Programme for IT in the NHS: Progress since 2006, HC (2007-2008) 484-I, para 13 Back

182   Luer connection systems are the standard way of making leak-proof connections for devices such as syringes, catheters and intravenous tubes. Back

183   Q 219 Back

184   Since giving evidence, Professor Toft has informed us that one company may soon be able to begin production (PS 83B). Back


 
previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2009
Prepared 3 July 2009