Human-Computer Interaction 3e Dix, Finlay, Abowd, Beale
Hospital da Trofa is a small privately run hospital near Porto in Northern Portugal. It was one of the partners in a European Commission funded project, Team-HOS. The Team-HOS project developed a participatory methodology aimed at the health sector and piloted in hospitals in three countries. The methodology focused on the stages prior to implementation which was contracted out.
At Trofa the project focused on the prescription and delivery of drugs and general bedside care. One of the key bottlenecks identified in the existing (non-IT) system was the communication between doctor, nurses and pharmacist and also the communication between nurses at shift change. The technical solution therefore focused on this triad. This was identified not simply by consultants observing current practice, but by a jointly owned process of clinical and administrative staff including the nurses, pharmacists and doctors who would be affected by the proposed system.
The system developed is focused on small hand-held PDAs. The hardware chosen incorporates a barcode reader and has 802.11 wireless networking (WiFi) as well as standard screen and stylus. A small number of wireless base stations means that the devices can be used throughout the ward at each bedside.
When doctors prescribe drugs these are entered through a PDA using drop down menus and the prescription is automatically entered into the patient record and the details made available to the pharmacist who can check for incompatibilities between drugs.
Standard PCs at the ward desk and also at the pharmacy give full access to the information system.
At the pharmacy every drug is wrapped in small barcoded packages, down to individual tablets. This is quite an overhead but when drugs are administered at the bedside, the barcode of the package is read. If there is any discrepancy between the treatment plan and the drug about to be administered the nurse is alerted. This prevents mistreatment or duplicate treatments.
In fact, nurses' training makes them highly meticulous, so the additional safety offered by the system would only rarely be significant, however even occasional errors in health care are important.
Perhaps as important as the clinical advantages are the commercial ones. Remember this is a private hospital and the additional checks of the barcoded drugs both give the patients a greater sense of security ... and also make the health insurance companies confident that they are only paying for drugs actually administered!
Recall (see Introduction, page 5) any system must be useful, usable and used. The commercial factors, and issues such as the sense of wellbeing given to patients, are critical to ensure actual adoption. Unlike many PDA-style hospital devices, this system is in real and successful use! Also in a healthcare setting feelings of security and safety are likely to lead to improved health - how we feel affects how we are!
The system does not only manage prescriptions, but complete patient care plans. As this nurse goes round the ward she can see at a glance what monitoring and treatments the patient requires. The small size also means that it does not intrude excessively between the nurse and patient, one of the most important relationships in the hospital.
Standard monitoring of blood pressure, pulse etc. are performed as usual and entered into the system. It would of course be possible to automatically capture much of this information, but this was not identified as a priority by the participatory process. The system was targeted at supporting not replacing nursing staff.
At shift change the system is used to help with the handover process. In the past nurses at shift change used to spend considerable time simply telling each other what treatments had been given, the condition of patients, etc. Whilst some of this is not captured within the system, it does relieve the nurses of the more mundane information transfer and has substantially reduced the time needed at shift change. It is thus acting as a form of asynchronous co-located groupware (see Chap 19, page 665) helping the nurses on one shift to communicate with those on the next.
At Trofa things went well because of a participatory approach that ensured that all stakeholders were committed to the implemented solution. However this is not always the case. The UsabilityNews article Nursing Studies show Weakness in Failing to Consult, tells how two healthcare applications foundered because they did not have support from nursing staff.
text - Alan Dix © 2004
photos - Team-Hos project © 2002