FOR HOW LONG IS DATA FROM PREVIOUS ADMISSIONS

ACCESSED BY HOSPITAL DOCTORS?

Ricardo João Cruz-Correia

Biostatistics and Medical Informatics Department, Faculty of Medicine,University of Porto

CINTESIS, Faculty of Medicine, University of Porto, Portugal

Altamiro Costa-Pereira

Biostatistics and Medical Informatics Department, Faculty of Medicine,University of Porto

CINTESIS, Faculty of Medicine, University of Porto, Portugal

Keywords: Electronic patient records, Data integration, Information use.

Abstract: Distinguishing relevant information enables for better user interfaces, as well as better storage management.

However, it is hard to distinguish between information really important to clinical care and only

occasionally desirable. We aim to answer for how long are clinical documents useful for health

professionals in a hospital environment considering its’ content and the context of information request. We

have studied the databases of a Virtual Electronic Patient Record that included (1) patient identification and

the list of clinical documents integrated, (2) the visualization logs; and (3) a hospital encounters database

that includes the list of encounters since 1993. Our results show that some clinical reports are still used after

one year regardless of the context in which they were created, although significant differences exist in

reports created in distinct encounter types. The half-life of reports by encounter type is 1.7 days for

emergency, 3.9 days for inpatient and 27.7 for outpatient encounters. We conclude that the usage of patients

past information (data from previous hospital encounters), varied significantly according to the setting of

healthcare and content.

1 INTRODUCTION

Patient records exist to memorize and communicate

the data existing on a particular individual, to help

deliver care to him or her. Records are not only an

information system but also a communication

system that enables communication between

different health professionals and between the ‘past

and present’ (Dick & Steen, 1997; Nygren, Wyatt, &

Wright, 1998).

Currently there are great quantities of stored data

regarding patients. Although great advances have

been made over the years (Cruz-Correia et al.,

2007), on-demand access to clinical information is

still inadequate in many settings, contributing to

duplication of effort, excess costs, adverse events,

and reduced efficiency (Feied et al., 2004). While it

is widely accepted that full access to integrated

electronic patient records and instant access to up-to-

date medical knowledge significantly reduces faulty

decision making resulting from lack of information

(Dick & Steen, 1997; Miller & Sim, 2004; Overhage

et al., 2002), there is still very little evidence that

life-long Electronic Health Records (EHR) improve

patient care (Clamp & Keen, 2007).

Distinguishing between relevant and useless

information enables for better user interfaces by

highlighting most relevant information, as well as

better storage management by choosing storage

devices with better performance for relevant data.

However, it is hard to understand what information

is really important to clinical care, and what is

simply occasionally desirable (Coiera, 1997).

Data age is usually one of the factors used to

assess importance, making new information more

relevant to the current search. But different data ages

differently according to its type, i.e., some clinical

reports describe situations less ephemeral than other

and so are found useful longer than others. Also, the

context of healthcare (e.g.: hospital environment,

219

ao Cruz-Correia R. and Costa-Pereira A. (2008).

FOR HOW LONG IS DATA FROM PREVIOUS ADMISSIONS ACCESSED BY HOSPITAL DOCTORS?.

In Proceedings of the First International Conference on Health Informatics, pages 219-222

 SciTePress

primary care, oncology) probably influences the way

information maintains its relevance.

We aim to study for how long are clinical

documents useful for health professionals in a

hospital environment.

2 BACKGROUND

In May 2003, the Department of Biostatistics and

Medical Informatics implemented a Virtual

Electronic Patient Record (HSJ-VEPR) (Cruz-

Correia et al., 2005) for the Hospital S. João (HSJ), a

university hospital with over 1 350 beds. The

system integrates clinical data from 12 legacy

departmental IS and the Diagnosis Related Groups

and Hospital Administrative databases, aiming to

deliver the maximum information possible to health

professionals. Over 700 medical doctors use the

system on a daily basis and the HSJ-VEPR retrieves

an average of 3000 new reports each day (in PDF or

HTML formats) (Cruz-Correia et al., 2005; Cruz-

Correia et al., 2006), adding up to 2 million reports

collected so far.

Each health professionals may access clinical

information either by reading the paper patient

record, using the HSJ-VEPR or using other IS

available on the hospital.

3 METHODS

3.1 Participants

This study is concentrated in the report

visualizations occurred in a two years period (2005

and 2006). In this period the hospital had 978 553

outpatients visits, 464 683 emergency visits and 82

444 inpatient visits. Reports’ half-life by feeder

system is analysis is based on the 3

quarter of 2006

view results.

3.2 Data Preparation

The data considered in this study existed in three

different Oracle schemas: (1) the HSJ-VEPR patient

database, which included patient identification and

the list of clinical documents integrated and; (2) the

visualization logs including sessions, health

professionals’ identification and category and

document views; (3) a hospital encounters database

that includes patient identification, the list of

encounters since 1993. These schemas use slightly

different patient identification numbers, so

transformation of these values was necessary to

create relations between the tables.

HSJ-VEPR system does not know in what

context (inpatient, outpatient or emergency) is the

user accessing each report. The context was induced

by confronting the date of view and the dates of the

different patient encounters. When the date of view

matches an encounter, that encounter is associated

with the visualization. When no match is made no

assumption is made regarding the encounter.

3.3 Clinical Report Half-life

Clinical reports’ percentile is calculated by grouping

all report views by type of encounter, ordering all

visualizations by date, and calculating its relative

position (current visualization position / number of

visualizations). This technique allows us to compare

the different encounter type groups by standardizing

the position of each view. Reports half-life refers to

the age of the report in percentile fifty.

4 RESULTS

Table 1 shows the number of visualizations taking in

consideration the context of report creation and the

context of report visualization. It should be noted that

more than a half of the reports seen in 2005/2006).

Table 1: Number and percentage of visualizations grouped by context of visualization and report creation in 2005 and 2006.

Concomitant

Previous encounter

Year Report viewed in

encounter

Emergency Inpatient Outpatient

Total

Emergency 861 40 334 16 511 24 447 21 2 153

Inpatient 18 929 62 4 794 16 3 337 11 3 352 11 30 412

Outpatient 154 1 1 158 4 5 150 18 22 043 77 28 505

2005

Total 19 944 33 6 286 10 8 998 15 25 842 42 61 070

Emergency 2 973 49 743 12 1 129 19 1 202 20 6 047

Inpatient 43 328 65 9 618 14 6 453 10 7 432 11 66 831

Outpatient 290 0 2 543 4 10 804 18 46 874 77 60 511

2006

Total 46 591 35 12 904 10 18 386 14 55 508 42 133 389

HEALTHINF 2008 - International Conference on Health Informatics

220

Figure 1: Reports half-life grouped by episode type according to views in 2005 and 2006.

In inpatient encounters more than 35% (38/35%)

report views regard previous encounters, without

any clear distinction of which are found more

relevant (previous emergency 16/14%, inpatient

11/10% and outpatient 11/11%).

In outpatient encounters almost all of the report

visualizations were of reports created in previous

encounters (99/100%). Most of them created in a

previous outpatient encounter (77/77%).

Figure 1 illustrates reports’ half-life by the type of

encounter when report was generated. It shows that

some clinical reports are still used after one year

regardless of the context in which they were created.

Nevertheless, outpatient reports are in average more

durable than inpatient reports and emergency reports.

The half-life of reports (percentile 50) by encounter

type is 1.7 days for emergency, 3.9 days for inpatient

encounters and 27.7 days for outpatient encounters.

Table 2 describes the reports’ half-life (median

of report age when viewed) group by department of

feeder system in the 3

quarter of 2006. It should be

noticed the great difference in reports’ half-life

regarding feeding system (e.g. half-life of the

pathology lab is 10 times greater than the clinical

pathology lab).

Table 2: Reports half-life (median of report age when

viewed) by department of feeder system in the 3rd quarter

of 2006.

Feeder system

Views

(n)

Half-life

(days)

Clinical Pathology 18 261 4.4

Imuno-hemotherapy 23 691 4.6

Obstetrics 241 8

Pneumology 457 15

Intensive Care 141 26

Gastroenterology 1 773 38

Gynaecology 100 44

Pathology 16 567 47

5 DISCUSSION

Our results show than many report visualizations

refer to previous encounters. Although the Hospital

has not a unique patient record (in paper or

electronic form), it is obvious that doctors which to

access to previous encounter reports. It is also

relevant that even older reports (more than one year)

are still found useful by doctors.

As more and more patient information is stored,

it is very important to efficiently select which one is

more useful and promote it in a scenario where the

scarceness of resources (screen space, disk space,

bandwidth and doctors’ time) is very real.

We intend to take in consideration reports’ half-

life in the next version of our system replacing the

first patient record screen, reports collected in the

last 24 hours, by a table in which the time interval is

different for each type of report. Outpatient reports

will be maintained in the list of last reports longer

than inpatient and emergency reports.

This study rises new questions regarding what

type of characteristics help maintain a report useful

over the years.

6 CONCLUSIONS

We conclude that the usage of patients past

information (data from previous hospital

encounters), varied significantly according to the

setting of healthcare and content.

FOR HOW LONG IS DATA FROM PREVIOUS ADMISSIONS ACCESSED BY HOSPITAL DOCTORS?

221

ACKNOWLEDGEMENTS

This study was supported by POCTI/0753/2004 -

Unit I&D CINTESIS nº 753/2004 from Fundação

para a Ciência e a Tecnologia.

REFERENCES

Clamp, S., & Keen, J. (2007). Electronic health records: Is

the evidence base any use? Medical Informatics and

the Internet in Medicine, 32(1), 5-10.

Coiera, E. (1997). Guide to medical informatics, the

internet and telemedicine: London: Chapman & Hall.

Cruz-Correia, R. J., Vieira-Marques, P., Costa, P.,

Ferreira, A., Oliveira-Palhares, E., Araujo, F., et al.

(2005). Integration of hospital data using agent

technologies - a case study. AI Communications, 18,

191-200.

Cruz-Correia, R. J., Vieira-Marques, P., Ferreira, A.,

Almeida, F., Wyatt, J., & Costa-Pereira, A. (2007).

Reviewing the integration of patient data: how systems

are evolving in practice to meet patient needs. BMC

Medical Informatics and Decision Making, 7(1), 14.

Cruz-Correia, R. J., Vieira-Marques, P., Ferreira, A.,

Oliveira-Palhares, E., Costa, P., & Costa-Pereira, A.

(2006). Monitoring the integration of hospital

information systems: how it may ensure and improve

the quality of data. Stud Health Technol Inform, 121,

176-182.

Dick, R., & Steen, E. (1997). The Computer-based Patient

Record: An Essential Technology for HealthCare.

Feied, C. F., Handler, J. A., Smith, M. S., Gillam, M.,

Kanhouwa, M., Rothenhaus, T., et al. (2004). Clinical

Information Systems: Instant Ubiquitous Clinical Data

for Error Reduction and Improved Clinical Outcomes.

Acad Emerg Med, 11(11), 1162.

Miller, R. H., & Sim, I. (2004). Physicians'use of

electronic medical records: barriers and solutions.

Health Aff (Millwood), 23(2), 116-126.

Nygren, E., Wyatt, J. C., & Wright, P. (1998). Helping

clinicians to find data and avoid delays. Lancet, 352,

1462-1466.

Overhage, J. M., Dexter, P. R., Perkins, S. M., Cordell, W.

H., McGoff, J., McGrath, R., et al. (2002). A

Randomized, Controlled Trial of Clinical Information

Shared From Another Institution. Ann Emerg Me,

39(1), 14-23.

HEALTHINF 2008 - International Conference on Health Informatics

222