Abstract

Introduction

Anecdotal evidence suggested that there was room for improvement in communications between our hospital (Ballarat Base Hospital) emergency department and local GPs. In particular, a 1995 survey showed that 84.4% of the Ballarat GPs who responded wished to be notified of their patients' admissions (Dr Mark Fitzgerald, Emergency Department Director, St John of God Hospital, personal communication). Ballarat Base Hospital medical officers were required to notify GPs of patient admissions by telephone. However, as 62% of patients are admitted from the emergency department (ED) between 1800 and 0800 (unpublished data) and GPs can be difficult to contact after hours, GPs were often not notified.

To rectify this problem, and to improve the relationship between the hospital and GPs, Ballarat Base Hospital and the Ballarat and District Division of General Practice aimed to establish a system of facsimile notification of GPs of patient admissions. Advantages of facsimile transmission of patient information include speed and accuracy of transfer, accessibility, low cost, and the possibility of direct transfer from the computer screen.^5-9 However, there are potential problems of misdirection of transmissions and of document security at the transmission destination.^10-14

In this study, we aimed, firstly, to examine the perceived success of this initiative in improving communications between the emergency department and GPs, and, secondly, to address potential problems and establish appropriate confidentiality protocols and controls governing the use of facsimile for this purpose.  

Methods

The ED purchased a dedicated facsimile machine with a preprogrammable number dialling facility to ensure that, if an incorrect number were pressed, the facsimile would be sent only to another participating GP. The machine was positioned in a secure area of the ED clerical office where incoming and outgoing facsimiles could not be accessed by unauthorised people. To test the system, a test facsimile was sent to all preprogrammed numbers of participating practices, which were asked to verify the security of their machines and their commitment to the confidentiality of the study by endorsing the test facsimile with the practice stamp and refaxing it to the ED.

Specially designed facsimile notification sheets recorded:

• The GP's name;

• Patient identification (name, age, date of birth, unit record number, address) and admission details (date, time, diagnosis, ward and inpatient unit);

• The patient's signature (indicating consent for their GPs to be notified and to provide any relevant medical information by facsimile); and

• A request from the ED for the GP to provide specific and any other relevant medical information in a space provided (the same facsimile could then be returned).

Facsimile notification of patient admission began on 6 June 1996; the study period ended on 31 July 1996.

Perceptions of communication between the ED and GPs before and after the establishment of facsimile notification were assessed by two questionnaires, posted to participating GPs in the week before and immediately after the study period. The questionnaires were identical, except for an additional question in the one sent at the end of the study period. For each question, respondents were asked to select the most appropriate response, and were invited to elaborate descriptively. GPs who did not respond were reminded by mail, fax or telephone.  

Statistical analysis

Results

During the study period, 887 patients were admitted from the ED. Facsimile notifications were sent to the GPs of 548 patients (61.8%). A further 74 patients (8.3%) were visitors or transferred from outside the area, 73 (8.2%) were unable to provide informed consent, 47 (5.3%) had GPs with no facsimile machine, 38 (4.3%) refused permission to send notification, 32 (3.6%) had no GP, and 14 (1.6%) had been admitted to the ward before authorising the notification. For the remaining 61 patients (6.9%), no explanation could be found for failure of notification. Usually, no attempt was made to notify GPs if a facsimile was not sent.

Of the 77 participating GPs, 44 (51.7%) completed the first, 55 (71.4%) completed the second, and 44 (57.1%) completed both the questionnaires. The demographic characteristics of the 77 GPs, comparing the 44 who completed the study with the 33 who did not, are shown in Box 1. Responses to the questionnaires, before and after the intervention, by the 44 GPs who completed the study are compared in Box 2.

After the intervention, there was a significant improvement in GPs' perception of overall communications from the ED (Question 1; P < 0.001). Similarly, GPs reported a significant improvement in the frequency of notification (Question 2; P < 0.001), and in communications (Question 3; P < 0.001) from the ED after one of their patients was admitted. Responses to Question 4 showed a significant change in GPs' preferred mode of communication after the intervention ( P < 0.01). Before the intervention, most preferred a direct telephone call from the admitting officer, while after the intervention most preferred a facsimile. When asked how notification of their patients' admissions from the ED would affect [had affected] their management of those patients (Question 5), most GPs stated, both before and after the intervention, that notification would encourage them to visit their patients in hospital and assist in planning postacute care. After recoding to adjust for the effect of multiple answers (more than one response was allowed), no statistically significant change was found as a result of the intervention.

The number of GPs responding that they had recently been aggrieved or inconvenienced as a result of the ED not notifying them of the admission of one of their patients (Question 6) fell significantly after the intervention ( P < 0.001). After the intervention, only two additional GPs responded that prompt notification of a patient's admission was of significant importance to the ongoing management of that patient (Question 7); this was not statistically significant. The number of GPs who had reservations about confidentiality associated with patients' medical records being transmitted by facsimile with their consent (Question 8) fell after the intervention ( P < 0.05).

Finally, the question included only in the second questionnaire showed that 43 GPs (97.7%) were prepared to support the continuation of the ED facsimile notification initiative. It also showed that 39 (88.6%) and 40 (90.9%) GPs, respectively, were prepared to support its use in all admissions and discharges at Ballarat Base Hospital.  

Discussion

Most GPs responded that notification allowed them to visit their patients in hospital, to contribute to inpatient care, to plan their patients' postacute care (e.g., home-help, meals-on-wheels, district nursing and family counselling), and to avoid interruption of the GP-patient relationship. Such continuity of care may affect readmission rates. Objective measurement of whether facsimile notification actually affects GP behaviour and patient outcomes may be worthy of further study.

A drawback of our study was its low response rate. Despite encouragement, a considerable number of GPs did not complete both questionnaires. While there were no obvious differences between those who did and did not complete the study, the low response rate may have introduced bias into the results by selecting for more motivated GPs, or those with a special interest in the study. The fact that more GPs responded to the second questionnaire may have indicated some enthusiasm for the project.

Lack of confidentiality has been identified as one of the most serious drawbacks of using facsimile machines in clinical practice. The sender loses control once a fax is transmitted, and information can be misdirected as a result of dialling wrong numbers.^{10,11,13,16,17}

None the less, we considered facsimile notification a more appropriate initial step than other accepted methods, such as encrypted electronic data transfer. The infrastructure was easy and relatively cheap to establish. It was also easy for the ED staff to use, with completion and transmission of each notification sheet taking approximately five minutes. Finally, the method allowed the patients to view the notification document before its transmission.

The safeguards we used in the facsimile transmission of patient med- ical information (Box 3) have been recommended by other investigators.^{6,11,12,14,16,17} We contend that, if these safeguards are established, the standard of care for the protection of patient information transmitted by facsimile should be at least as high as that provided by conventional mail. A message reaching a secure facsimile machine is no more susceptible to loss, misdirection or unauthorised access than an opened letter in a doctor's "in-tray".

It has been suggested that an authorisation procedure should be in place to facilitate obtaining informed consent for any facsimile transmission of personal information.^{6,12,14,16,17} We felt that if patients were shown the information that was to be transmitted, its destination and the format in which it was to be sent they could make an informed decision about whether or not GP notification was appropriate.

The number of patients who refused consent for facsimile notification (38, or 4.3%) was surprising; our study had not been designed to record the reasons for these refusals and this matter deserves further study. Regardless of the reasons for refusal, patients could suffer distress if medical information was transmitted without consent, and legal action could result from misdirection of unauthorised, non-urgent medical information.

Our protocol represented a change in Ballarat Base Hospital's facsimile policy as consent had previously been assumed. In most hospitals, providing the name of the GP at registration or admission is considered implied consent for the hospital to communicate directly with the GP. Letters, discharge summaries and death notifications may be sent by facsimile, electronic transfer or mail. With increasing use of telecommunications and information technology, we contend that it may be appropriate to question some of the practices that are currently undertaken routinely with only the implied consent of patients.

Other Australian emergency departments are developing facsimile and electronic data systems for the transfer of patient information. Facsimile remains a "paper-based" technology and is likely to become superseded by electronic data transfer systems. Indeed, the American College of Emergency Physicians believes that, along with facsimile, electronic data facilities should be available to all emergency departments.⁶ Various electronic data transfer systems have been developed to coordinate the shared care of patients,¹⁸ to establish electronic medical records,¹⁹ to facilitate communication between hospitals and GPs,²⁰ and to streamline patient referrals.²¹ While these "paperless" systems may be superior to facsimile, they also raise confidentiality issues. These are being addressed with the use of cryptography,²² digital signatures,²⁰ "need to know" staff clearances,²⁰ and legislation.^20,23 Meanwhile, facsimile is accessible and relatively inexpensive, and is likely to remain useful, at least in the near future, until electronic systems become more prevalent.

We recommend the use of facsimile transmission for notification of patient admission, provided that adequate safeguards are in place, and that informed consent is obtained before transmitting medical information.

Following the success of facsimile notification of admissions from the ED, Ballarat Base Hospital has continued its use and is considering extending it to all hospital admissions and discharges, ED discharge letters and, possibly, to service providers other than GPs. Presently, the hospital is investigating software which will allow computer-generated facsimile transmission and is looking ahead to electronic data transfer.  

References

1. Bella JI, Jamieson WE. Improving the continuity of care between general practitioners and public hospitals. Med J Aust 1994; 161: 656-659.
2. Freeman G. Continuity of care in general practice: a review and critique. Fam Practitioner 1984; 1: 245-252.
3. Interaction with Fremantle Hospital: final report. Fremantle: Fremantle Regional Division of General Practice, 1994: 1-40.
4. Morrison WG, Pennycook AG, Makower RM, Swann IJ. The general practitioner's use and expectations of an accident and emergency department. J R Soc Med 1990; 83: 237-240.
5. Magennis AW. Fax units in general practice. Aust Fam Physician 1989; 18: 1259-1264.
6. American College of Emergency Physicians. The use of facsimile machines and electronic data transfer in the emergency department [policy statement]. Ann Emerg Med 1993; 22: 266.
7. Spigelman A. Faxed electronic summaries are valued by general practitioners [letter]. BMJ 1995; 311: 746-747.
8. Cole DR, Johnson MS, Heaton CJ, Petti M. Fax/modem board communications decrease preceptor communication costs. Fam Med 1994; 26: 418-420.
9. Yamamoto LG, Wiebe RA. Improving medical communication with facsimile (fax) transmission. Am J Emerg Med 1989; 7: 203-208.
10. Marr P. Maintaining patient confidentiality in an electronic world. Int J Biomed Comput 1994; 35 Suppl: 213-217.
11. Brent N. Facsimile systems revised: focus on confidentiality and privacy. Home Healthcare Nurse 1991; 9: 6-8.
12. Larkin GL, Moskop J, Sanders A, Derse A. The emergency physician and patient confidentiality: a review. Ann Emerg Med 1994; 24: 1161-1167.
13. Carman D, Britten N. Confidentiality of medical records: the patient's perspective. Br J Gen Pract 1995; 45: 485-488.
14. Capen K. Facts about the fax: MDs advised to be cautious. Can Med Assoc J 1995; 153: 1152-1153.
15. SPSS/PC+ statistics [computer program]. Version 4.0. Chicago: SPSS Inc., 1990.
16. Genesen LB, Sharp HM, Genesen MC. Faxing medical records: another threat to confidentiality in medicine [letter]. JAMA 1994; 271: 1401-1402.
17. Grant AE. Legal matters -- facsimile transmissions. Canadian Nurse 1996; 92: 47.
18. Branger P, van't Hooft A, van der Wouden HC. Coordinating shared care using electronic data interchange. Medinfo 1995; 8: 1669-1674.
19. Walker D. Transferring electronic medical records. Aust Family Physician 1997; 26: 48-55.
20. Fisher F, Badge B. Data security and patient confidentiality: the manager's role. Int J Bio-Med Comp 1996; 43: 115-119.
21. Gaudet LA. Electronic referrals and data sharing: can it work for health care and social service providers? J Case Management 1996; 5: 72-77.
22. Biskup J, Bleumer G. Cryptographic protection of health information: cost and benefit. Int J Bio-Med Comp 1996; 43: 61-67.
23. Cassidy SO, Sepulveda MJ. Health information privacy reform. J Occup Environ Med 1995; 37: 605-614.

Home | Issues | eMJA shop | Terms of use | Classifieds | More... | Contact | Topics | Search

The Medical Journal of Australia    eMJA