Abstract
Research on the use of information technology in healthcare has focused on hospitals and Health Management Organizations (HMOs). However, little has been done to study the use of IT in group medical practices. In 1994, we conducted a pilot study of group medical practices and then repeated this pilot study in 2003 to obtain a longitudinal picture of the IT services used by these private practices. Researchers can use this to form ideas of the important issues and changes involved in IT usage in group medical practices over the past decade thus providing a needed benchmark to fill a gap in the existing literature and that can be used to compare domestic as well as international practices. For example, an expanded form of this study will be conducted in Taiwan this summer.
Introduction
Studies of the introduction of computer technology in medical settings have focused on hospitals (Griffith & Sobol, 2000; Sobol, Humphrey, & Jones, 1992; Sobol & Smith, 2001) and more recently on HMOs. In these studies, such issues as barriers to the introduction of technology in hospitals, returns to adoption of technology, and the market status of the adoption of different technologies have been studied. It was found that there are many barriers to the adoption in hospitals. The longitudinal issues of what the changes have been in the last decade have been studied with the results that certainly there has been an increase in adoptions over the past decade. These increases have occurred in both transactional, informational, and strategic uses of technology (Sobol & Woods, 2000). This trend is expected to increase. A survey in 2002 by Sheldon I. Doren-fest & Associates of Chicago indicated that IT spending on healthcare in 2002 would be $21.6 billion (Dorenfest, 2002).
While the focus has been hospitals and HMOs, very little has been done to study the use of IT in group medical practices both small and large. This is the case even though researchers have for years trumpeted the impact of IT on physician’s practice (Rodger, Pendharkar, & Paper, 1996; Shine, 1996). In 1994, we conducted an initial study of group medical practices of three or more doctors; we completed a later study in 2003 to obtain a longitudinal picture of the IT services used by these private practices. While this is not yet the definitive study of IT in group medical practices, it can be used to form ideas of the important issues and changes involved in IT usage in the smaller group medical practices over the past decade thus providing a needed benchmark to fill a gap in the existing literature and to start an intensive study of changes in IT usage.
In this article, we look at the differences in computer usage, computer facilities, sources of computer information, and the satisfaction with computer usage in group medical practices from 1994 to 2003. We compare these characteristics and the amount of time spent on business issues by size of practice and years in practice for group medical practices studied in 1994 and 2003.
background
There has been a great deal of research on IT as well as healthcare. Unfortunately, much of this has been of limited use to practicing physicians. From the computer science side of research, most work has been done on theoretical computing structures. This includes work such as neural net applications of drug/plasma levels (Tolle, Chen, & Chow, 2000) or parsing methods for biomedical texts (Leroy, Chen, & Martinez, 2003). When trying to overlap IT and near term healthcare concerns, the research has tended to focus on public policy (Magruder, Burke, Hann, & Ludovic, 2005) or on hospitals. This includes work done on hospitals and adoption of computer-based IT (Sobol et al., 1992; Sobol & Woods, 2000), as well as the impact of IT use on hospital staffing and payroll (Sobol & Smith, 2001). Other work has focused on the barriers to IT adoption within healthcare (Economist, 2005; Sobol, Alverson, & Lei, 1999).
On the other end of the spectrum, some research has been conducted on issues surrounding IT in private practices. This research has tended to be very specific in nature, however. This includes whether or not medical practices should hire an IT person or outsource (Lowes, 2005) or the use of electronic billing systems by private practices (Burt, 2005). Other research tends to focus on a hot technology that is currently being embraced such as electronic medical records (Miller & Sim, 2004; Palattao, 2004). What is lacking is an overall benchmark or “snapshot” of overall IT use by private practice physicians. That is the goal of this research.
methodology
In the summer of 1994, a mail survey was sent to a sample of 270 multiple physician groups within Maricopa County, Arizona, who were chosen from lists of a value-added reseller. These practices were medical groups containing three or more physicians. A total of 65, or a response rate of 24%, of usable replies were received. This is a good response for a mail survey. In the summer of 2003, 54 physicians were surveyed in group practices of three or more in the Arlington/Mansfield area of Tarrant County, Texas. The surveys were given to the business managers’ offices and were returned by mail. Thus, in essence both were mail surveys utilizing the same questionnaire. Both counties (Maricopa and Tarrant) included large metropolitan areas (Phoenix and Fort Worth), were 71 to 77% white and had median per capita income of approximately $22,250 to $22,500. In both counties, 65 to 69% of the people were in the labor force. These statistics show both interview sites to be similar Southwestern areas. The group practices in Tarrant County were randomly chosen from a list of 525 physicians in group practices and provided a sampling rate of 10%. The size of the practices and the years in practice in both surveys were well-distributed over a wide spectrum of sizes and age of practice. In the following paragraphs, we will look at the use of computers in the group medical practice with the primary perspective of assessing how this usage has changed over the last decade.
First, we will use rank order correlation to determine if the orders of importance for various technologies (hardware and applications) have remained the same over the decade. Spearmans’ rho was chosen for this because it makes no assumptions about the shape of the relationship between variables. Secondly, we will employ tests of the differences in proportions for software and hardware adoptions, important types of technology and applications, time spent on business applications, and sources of computer information, comparing 1994 and 2003. These tests will be developed for whole groups and will also focus on differences by the size of the practice (number of doctors) and the number of years in practice.
survey results
characteristics of the samples
The practices in the 1994 sample tended to be larger in terms of the number of physicians than the 2003 sample. As we can see from Table 1A, about one quarter of the group practices in 1994 were less than five while 60.4% of those in 2003 were less than five. If we look at practices of less than 10, 50% of the 1994 sample can be compared to 75% of the 2003 group. Why this disparity? There are many different factors. For example, the Arlington area is a fast-growing area and new start-up practices are forming. However, these are not just new physicians but include older practices that are moving to the faster-growing areas. In addition, during the past 10 years, many practices have become part of hospital-based groups in order to save on costs. This leveraging of resources among a large hospital network has meant that there is less of a need to have larger individual practices in order to distribute costs. There may be other factors, but these are hard to delineate. Because of this, for the studies in this paper, we will look at the variables by practice size so we can study small vs. large practices.
Table 1. Characteristics of samples, 1994 vs. 2003
|
A. Number of Physicians In Practice |
1994 |
2003 |
|
Less than 5 |
24.1% |
60.4% |
|
5 but < 10 |
25.9 |
15.7 |
|
10 but < 25 |
31.6 |
15.7 |
|
25 but < 50 |
18.4 |
8.2 |
|
|
100% |
100% |
|
B. Years in Practice |
|
|
|
0-5 |
23.1% |
17.3% |
|
6-10 |
16.9 |
11.5 |
|
11-15 |
18.5 |
17.3 |
|
16-20 |
12.3 |
17.3 |
|
21+ |
27.7 |
36.6 |
|
No answer |
1.5 |
0.0 |
|
|
100% |
100% |
Table 2. Current computer systems, 1994 vs. 2003
|
System |
1994 |
2003 |
|
Individual PCs (not connected) |
15.4% |
17.5% |
|
Network connected PCs |
35.4 |
61.4* |
|
Midrange or Mini Computer |
16.9 |
8.8 |
|
Mainframe Computer |
13.8 |
3.5* |
|
Network PCs & Mainframe |
3.1** |
– |
|
Network PCs & Mini |
4.6** |
– |
|
Service Bureau |
– |
8.8** |
|
Don’t Know |
10.8 |
– |
|
|
100% |
100% |
* Indicates that the differences are significant at the 0.05 level using a paired t-test, two tail. ** Reflects differences in available services.
In terms of years in practice, the samples are similar. In Table 1B, we see that in 1994, about 23.1% of the practices had been in business five years or less, while 17.3% of the 2003 practices had been in practice 5 years or less. If we look at 6 to 10 years in practice, 16.9% of the 1994 group as opposed to 11.5% had been in practice 6 to 10 years. Adding these groups to form groups in practice 0 to 10 years we get 40% for 1994 and 27.8% for 2003. Thus, the 2003 practices tended to be smaller and somewhat older. To account for these differences, we will separate a number of our analyses by practice size and years in practice.
current computer systems
The computer systems used have changed over the last decade. We can see from Table 2 that in 1994, 15.4% used individual PCs (non-connected, in even a local area network) while 17.5% in 2003 used individual PCs (non-connected). This remained about the same. However, there has been a big move to network-connected PCs.
Types of Business Applications used
We have examined the types of computer equipment in group practice offices; we now turn to the types of applications used by these offices. This is important because when the doctors were asked if they were satisfied that the applications used met their business requirements, 86.7% of doctors in 1993 were. In 2003, the level of satisfaction had risen slightly to 91.0%. So while there have been complaints that physicians are not open to using information technology in hospital settings (Florien, 2003), physicians apparently are satisfied using IT in their own practices. Table 3 indicates the types of uses of computer applications for the 1994 and 2003 samples. The rank order of overall uses in 2003 is highly correlated with uses in 1994.
Since the earlier sample contained a larger number of larger medical practices, we will separate the study by group size to make the statistics more comparable (see Table 4). Generally, the order of importance of the usage of applications has remained the same (r = 0.949). This should come as a surprise since the basic business aspects of managing a business/practice have not changed. In 1995, a survey of physician managers found that the key issues they needed support in were personnel management, computing, budgeting, and financial management issues (Cordes, Rea, Rea, & Vuturo, 1995).
Table 3. Applications used in group medical practice, 1994 vs. 2003
|
Percent Who Used Application Applications |
1994 |
2003 |
|
Personnel Scheduling |
48.3% |
26.9%* |
|
Facility Scheduling |
21.7 |
28.8 |
|
Patient Scheduling |
51.7 |
55.8 |
|
Insurance Billing |
95.0 |
86.5 |
|
Practice Billing |
85.0 |
80.8 |
|
Business Record Keeping |
78.3 |
65.4 |
|
Patient Record Keeping |
53.8 |
46.7 |
|
Using Networking Software |
51.9 |
30.0*** |
|
Using Hospital Network Software |
32.7 |
11.7*** |
|
Vendor Networking |
10.0 |
9.6 |
|
Expert Systems |
10.0 |
11.5 |
|
Imaging Technology |
– |
11.5** |
|
Voice Recognition |
0 |
7.7** |
Spearman’s Rank Order Correlation Coefficient rho = .949, p = .00001
* Significant at 0.05 level.
** Reflects differences in available services.
*** Reflects the fact that practices were automatically networked.
t There were at least three possible ways to compare the percentages: Spearmans’ rho, Kendall s tau and gamma. Since the questions were the same but the samples were not paired, Spearman s rho was chosen.
Table 4. Application used by group practice size, 1994 vs. 2003
|
Applications |
Groups Less Than 10 |
Groups 10 or Greater |
|||
|
|
1993 |
|
2004 |
1993 |
2004 |
|
Personnel Scheduling |
42.3% |
|
13.3% |
48.0% |
32.4% |
|
Facility Scheduling |
15.4 |
|
33.3 |
28.0 |
27.0 |
|
Patient Scheduling |
53.8 |
|
53.3 |
50.0 |
56.8 |
|
Insurance Billings |
100.0 |
|
100.0 |
92.0 |
81.1 |
|
Practice Billing |
84.6 |
|
73.3 |
80.0 |
83.8 |
|
Business Record Keeping |
80.8 |
|
53.3 |
80.0 |
70.3 |
|
Patient Record Keeping |
34.6 |
|
66.7 |
52.0 |
48.6 |
|
Practice Networking |
19.2 |
|
33.3 |
36.0 |
59.5 |
|
Hospital Networking |
3.8 |
|
46.7 |
24.0 |
27.0 |
|
Vendor Networking |
7.7 |
|
6.7 |
12.0 |
10.8 |
|
Use of Expert System |
3.8 |
|
20.0 |
16.0 |
8.1 |
|
Imaging Technology |
3.8 |
|
13.3 |
28.0 |
10.8 |
|
Voice Recognition |
0 |
|
6.7 |
0 |
0 |
|
|
rho p = |
= .767 .002 |
|
rho = .925 p = .00006 |
|
Practice networking is higher for large groups and seems to have risen for all groups since 1994. Hospital networking by computer for groups of less than 10 physicians has risen considerably from 3.8% to 46.7%. Vendor networking seems to have stayed the same.
Table 5. Applications used by years in practice 1994 vs. 2003
|
Applications |
|
|
Years in Practice |
|
|
|
|
0 |
- 10 |
11 – 20 |
21 Plus |
|
|
|
1994 |
2003 |
1994 2003 |
1994 |
2003 |
|
Personnel Scheduling |
71.4% |
13.3% |
38.9% 38.9% |
35.3% |
26.3% |
|
Facility Scheduling |
33.3 |
33.3 |
11.1 33.4 |
17.6 |
21.1 |
|
Patient Scheduling |
60.9 |
53.3 |
50.0 61.1 |
35.3 |
52.6 |
|
Insurance Billings |
95.8 |
100.0 |
94.4 83.4 |
94.1 |
78.9 |
|
Practice Billing |
87.5 |
73.3 |
77.8 83.4 |
88.2 |
84.2 |
|
Business Record Keeping |
83.3 |
53.3 |
66.7 88.9 |
77.8 |
52.6 |
|
Patient Record Keeping |
54.2 |
66.6 |
27.8 66.7 |
42.9 |
31.6 |
|
Practice Networking |
37.5 |
40.0 |
27.8 72.3 |
17.6 |
47.4 |
|
Hospital Networking |
8.3 |
46.7 |
11.1 38.9 |
17.7 |
15.8 |
|
Vendor Networking |
12.5 |
6.7 |
11.1 11.1 |
5.8 |
10.5 |
|
Use of Expert Systems |
0 |
20.0 |
16.7 11.1 |
17.7 |
10.5 |
|
Imaging Technology |
16.7 |
13.3 |
16.7 5.6 |
17.6 |
15.8 |
|
Voice Recognition |
4.2 |
0 |
0 16.7 |
0 |
5.3 |
|
|
rho |
= .719 |
rho = .771 |
rho = .828 |
|
|
|
p = |
00561 |
p = .003 p |
= .00048 |
We now consider the relationship between years in practice and usage of different computer applications (Table 5). We used rank order correlation to compare the relative amount of use of each of these applications by years in practice (Table 5), and the rankings have remained similar over the decade. However, different applications show differing amounts of use.
Importance of Different Types of savings with Respect to Business Aspects of the Medical practice
We then looked at the overall rankings of the importance of different types of savings or improvements to business practice that could be achieved by utilizing computers (Columns 5 & 6, Table 6). We asked respondents to rank the importance of different types of savings from computer adoptions on a scale from 1 to 5 (with 5 being very important). In more recent years (2003), respondents have been inclined to say that different savings are very important. On every one of 13 categories in Table 6, columns 5 and 6, the 2003 sample is 15% to 35% higher than their counterparts were in 1994. In both years, improving insurance claim processing was ranked as most important. In 1994, 70.2% said this, but by 2003, 85.8% found this type of savings most important. Other important savings issues in 2003, with 70% or more ranking them as important, were improving patient record keeping, access to patient or hospital information, increasing cash flow, and reducing administrative overhead.
The next area was the comparison of the importance of different types of savings by the age of the practice (see Table 7). These could be cost reduction or quality improvements. In Table 6, we classified the savings. Overall, the three groups (0 to 10, 11 to 20, 21+) order of importance (by use) of the applications was the same although the individual usage percentages differed widely.
Time spent on Business Aspects of practice
A very striking difference in the decade was the increased amount of time being spent by doctors on the business aspects of their practices. In 1994, 64.6% of the doctors said that they spent less than 10% of their time on business aspects, while in 2003, only 23.1% spent less than 10% on business, indeed they spent far more time on these aspects. About a quarter of the doctors in 1994 said they spent up to 25% of their time on business, and 21.2% of the 2003 sample recorded this answer. The striking difference was that, in 2003, 55.7% of the doctors spent more than 25% of their time on business as compared to only 9.2% of the doctors in 1993, reporting that they spent more than 25% of their time on business issues (see Table 8).
Table 6. Importance of different types of savings from computer adoptions, 1994 vs. 2003 by a number of people in the group
|
|
|
Less than 10 |
10 or |
More |
All |
|
|
|
|
|
1994 |
2003 |
1994 |
2003 |
1994 |
2003 |
|
CR |
Reducing administrative overhead |
59.3% |
66.7% |
44.0% |
80.6% |
54.0% |
70.3% |
|
CR |
Improving insurance claim processing |
70.4 |
86.7 |
68.0 |
83.3 |
70.2 |
85.8 |
|
QI |
Improving patient record keeping |
37.0 |
80.0 |
52.0 |
77.8 |
45.9 |
79.4 |
|
QI |
Access to patient or hospital information |
26.9 |
80.0 |
52.0 |
63.9 |
38.4 |
75.8 |
|
CR |
Increasing cash flow |
55.6 |
66.7 |
52.0 |
88.9 |
55.9 |
72.5 |
|
QI |
Use of computer technology |
22.2 |
66.7 |
33.3 |
66.7 |
23.4 |
66.7 |
|
QI |
Use of image storage of patient records |
14.8 |
60.0 |
8.0 |
52.8 |
13.3 |
58.1 |
|
CR |
Reduce service bureau costs |
28.0 |
33.3 |
21.4 |
50.8 |
24.5 |
37.8 |
|
QI |
Enhance professional image |
37.0 |
46.7 |
40.0 |
77.8 |
39.3 |
54.8 |
|
QI |
Business training for support or staff |
11.1 |
46.7 |
24.0 |
61.1 |
19.6 |
50.4 |
|
QI |
Business management of practice |
33.3 |
66.7 |
37.5 |
66.7 |
35.0 |
66.7 |
|
CR |
Financial performance/ controls |
44.4 |
66.7 |
52.0 |
72.2 |
47.5 |
68.1 |
|
QI |
Business planning |
22.2 |
60.0 |
36.0 |
69.4 |
27.5 |
62.4 |
rho = .440 rho = .775 rho = .746
p = .133 p = .002 p = .003
CR = cost reduction QI = quality improvements
Table 9 shows the importance of different types of savings by time spent on business aspects. One of the key changes from 1994 to 2003 was that, in 1994, both groups sited insurance claims processing as most important. On the other hand, in 2003, access to patient and hospital information had become most important thus reflecting a key shift in focus.
Table 7. Importance of different types of savings from computer adoptions, 1994 vs. 2003 by years in practice
|
|
0 |
10 Years |
11 |
20 |
|
21+ |
|
|
1994 |
2003 |
1994 |
2003 |
1994 |
2003 |
|
Reducing administrative overhead |
54.0% |
66.7% |
70.6% |
88.2% |
37.5% |
73.7% |
|
Improving insurance claim processing |
66.7 |
88.7 |
76.5 |
88.9 |
68.8 |
78.9 |
|
Improving patient record keeping |
42.3 |
80.0 |
47.4 |
94.5 |
41.2 |
63.2 |
|
Access to patient or hospital information |
28.0 |
80.0 |
31.6 |
82.7 |
58.9 |
47.4 |
|
Increasing cash flow |
48.0 |
66.7 |
52.6 |
88.9 |
70.6 |
89.5 |
|
Use of computer technology |
12.0 |
66.6 |
26.3 |
76.4 |
41.2 |
57.9 |
|
Use of image storage of patient records |
8.0 |
60.0 |
15.8 |
58.4 |
17.7 |
47.4 |
|
Reduce service bureau costs |
13.0 |
33.3 |
25.0 |
47.2 |
41.2 |
52.6 |
|
Enhance professional image |
32.0 |
46.7 |
31.6 |
82.7 |
52.9 |
73.7 |
|
Business training for support or staff |
12.0 |
46.7 |
21.1 |
58.4 |
29.4 |
63.2 |
|
Business management of practice |
25.0 |
66.7 |
31.6 |
53.5 |
47.1 |
78.9 |
|
Financial performance/ controls |
40.0 |
66.7 |
47.4 |
66.0 |
58.9 |
78.9 |
|
Business planning |
16.0 |
60.0 |
36.8 |
66.7 |
35.3 |
73.7 |
rho = .688 rho = .672 rho = .554
p = .009 p = .004 p = .049
Table 8. Percent of doctor’s time spent on the business aspects oftheirpractice, 1994 vs. 2003
|
Percent of Time Spent in Business Aspects of Practice |
1994 |
2003 |
|
Less than 10% |
64.6% |
23.1* |
|
10% up to 25% |
24.6 |
21.2 |
|
More than 25% |
9.2 |
55.7* |
|
No Answer |
1.6 |
– |
|
|
100% |
100% |
Table 9. High importance of different types of savings from computer adoptions, 1994 vs. 2003 by time spent on business aspects
|
Time Spent by Doctors on Busin Less Than 10 Percent |
Aspects of Their Practic More Than 10% |
|||
|
1994 |
2003 |
1994 |
2003 |
|
|
Reducing administrative overhead |
53.8% |
16.7% |
54.5% |
27.5% |
|
Improving insurance claim processing |
64.1 |
25.0 |
81.8 |
27.9 |
|
Improving patient record keeping |
46.1 |
50.0 |
45.5 |
64.9 |
|
Access to patient or hospital information |
39.5 |
100.0 |
36.4 |
82.7 |
|
Increasing cash flow |
59.0 |
83.3 |
50.0 |
80.2 |
|
Use of computer technology |
23.8 |
50.0 |
22.7 |
70.3 |
|
Use of image storage of patient records |
17.8 |
41.7 |
4.5 |
55.0 |
|
Reduce service bureau costs |
31.5 |
41.7 |
11.1 |
55.0 |
|
Enhance professional image |
38.5 |
41.7 |
40.9 |
35.0 |
|
Business training for support or staff |
17.9 |
8.3 |
22.7 |
7.6 |
|
Business management of practice |
34.2 |
0.0 |
36.4 |
12.6 |
|
Financial performance/controls |
46.2 |
8.3 |
50.0 |
7.6 |
|
Business planning |
20.5 |
8.3 |
40.9 |
9.9 |
rho = .198 rho = -.163
p = .518 p = .594
Correlating 1994 (< 10% vs. 10%+), rho = 0.884, p = 0.000006 Correlating2003 (< 10% vs. 10%+), rho = 0.957, p = 0.0000003
Sources of Information on Running Business Aspects of Their practice
The doctors in both surveys were asked, “Where do you get the information you need to run the business aspects of your practice?” This question suggested sources and allowed doctors to fill in blank lines. Also, it allowed respondents to check many answers. We will compare overall answers for the two surveys (Table 10). The ranking stayed approximately the same for 2003 as it had been in 1994 (rho = 0.813, p = 0.04). However in 2003, there was more reliance on associations and professional meetings and on professional journals. Use of consultants had tapered off, but vendors were more likely to be relied upon for information.
discussion
In viewing Table 2, we see that, in 1994, 35.4% of the offices had network-connected PCs while, in 2003, 61.4% had network-connected PCs. In 1994, 16.9% of the offices had mid-range or mini-computers while, in 2003, only 8.8% had these large computers. No doubt this is due to the enhanced capacities of today’s PCs. No one in 2003 reported a network of PCs and mainframes, while, in 1994, about 7.7% used these facilities. The past decade has also shown a difference between the computing services that are available for physicians to use. For example, in 2003, 8.8% of the practices reported that they used outside service bureaus to handle their computer needs. Application Service Providers (ASPs) were not commonly available in 1994.
Table 10. Sources of information to run business
|
Percent Who Checked Answers |
|
|
|
Sources |
1994 |
2003 |
|
Colleagues |
55.4% |
53.1% |
|
Associations/Professional Meetings |
46.2 |
67.3* |
|
Professional Journals |
46.2 |
59.2* |
|
Business Publications/Newspapers |
7.7 |
10.2 |
|
Consultants |
55.4 |
34.7* |
|
Universities |
3.1 |
14.3 |
|
Vendors |
21.5 |
38.8* |
|
Other (CPAs, Billing Companies, |
|
|
|
Business Managers) |
13.8 |
10.2 |
rho = .813 p = .040
Table 3 shows that, in general, the rank orders for application usage by practice size has remained the same in the last decade (rho = 0.767 and 0.925 respectively). Personnel scheduling by computer is done more often on the computer by large firms, but has gone down over the decade. Facility scheduling has gone up and patient scheduling has stayed the same or increased slightly. It seems that insurance billings for large firms using computers have decreased somewhat but for smaller firms 100% of insurance billing is done by computer. This difference may be due to age differences of physicians (older physicians tending to be in larger groups). We will assess the impact of age on IT usage later.
Looking at Table 4, we see the use of expert systems has gone up considerably in small practices from 3.8% to 20%. This is due to the fact that decision support systems have played an increasingly widespread role as the healthcare industry has embraced managed care (Dutta & Heda, 2000). This was predicted by McCauley and Ala (1992). One other area that has seen change is that of imaging technology. This may be due to the impact of telemedicine. In the early 1990s, imaging technology was very expensive, and only large groups and hospitals could afford it. Today, there is a minimal cost associated with it and new applications such as remote or telemedicine has attracted new physicians (Prater & Roth, 2003).
If we look at Table 5, we see that different applications have grown and fallen in importance depending on the age of the practice. The younger practices (0 to 10 years in practice) were more likely to use hospital networking, patient record keeping, and expert systems than their counterparts were in 1994 and 2003. They were less likely to use the computer for personnel scheduling, probably because they were in smaller practices. If we look at the middle group, 11 to 20 years in practice, they were more likely to do patient scheduling, patient record keeping, and practice networking than were their counterparts in 1994. So relatively new (10 years or less) and intermediate-aged practices (11 to 20 years) were more likely to use computers than the older practices (21+ years). In most other areas, the applications listed were utilized by the same percentage or less for each of the years in practice groups in 2003 as compared to 1994.
Table 6 shows that except for access to patient or hospital information, the same issues were very important in 1994 as in 2003 but not at the 70% level. No doubt because there was relatively little networking capacity in 1994, this issue was only ranked as very important by 38.4% of the sample. Assessing the importance of savings with relation to practice size (Table 6, Columns 1-4), the three most important issues for smaller practices (less than 10 physicians) in 2003 are improving insurance claims processing, improving patient record keeping, and access to patient hospital information. In 1994, the smaller practices chose improving insurance claims processing, increasing cash flow, and financial performance controls. So they were stressing the importance of the computer for billing and financial purposes. The smaller practices were not as likely to rank order these types of savings in the same orders in 1994 as they did in 2003 (rho = 0.44, p = 0.133). For the larger practices (10 or more physicians) in 2003, the most important savings were increasing cash flow, improving insurance claims processing, and reducing administrative overhead. For 1994, the top importance (for large practices) was for improving insurance claims processing. There were four other savings that tied for second place, improving patient record keeping, access to patient and hospital information, increasing cash flow, and financial performance controls. Uses devoted to quality improvements (QI) almost doubled in importance from 1994 to 2003. On the other hand, uses aimed at cost reduction increased in importance by 15 to 20 percentage points. Thus, it seems that early computer introductions were primarily devoted to cost reductions while later introductions may be more focused on quality improvements such as enhancing professional image, access to patient hospital information, and improving patient record keeping. Looking at the younger practices (0 to 10 years), we see that here again improving insurance claims processing was the most important issue in 1994 and 2003. Reducing administrative overhead was next in importance in 1994 and then increasing cash flow ranked third. In 2003, improved patient record keeping and access to patient and hospital information tied for second place. Thus in 2003, the informational and strategic uses of IT (such as long-term planning and marketing improvement) are becoming as important as transactional uses. However, the top uses were mainly those that led to cost reductions rather than quality improvements.
Using Table 7, and looking at the mid-term practices (11 to 20 years), again insurance claims processing was most important in both years, reduction of administrative overhead came in second in 1994 and third in 2003. Tying for first in 2003 was increased cash flow, and improving insurance claim processing. Tying for third in 2003 was a new variable enhanced professional image, and access to patient or hospital information. Here, we see the older doctors starting to focus on quality improvements. Finally, when we look at the older practices (21 years or more), increased cash flow comes out first for 2003 with improving claims processing, business management of the practice and business training for support or staff and financial controls vying for second place. Evidently, the older firms were sensing the need for updates and for more training and better business management practices. In 1994, the older practices ranked increasing cash flow first and improving insurance claims processing second. So over the years, the older practices have sensed the need for better business management and quality improvements. This need has probably been instrumental in the encouragement of the recent developments in management training and executive education for doctors and dentists (Glasser, 1997; Lazarus, 1999; Lipson, 1997).
With respect to the data from Table 8, we see that in 1994, 65% of the doctors spent less than 10% of their time of the business aspects of their practice. By 2003, 55.7% of the doctors reported that they spent more than 25% of their time on business aspects. With all of the different current issues such as HMOs, health insurance, Medicare, Medicaid, and so forth, doctors must personally make more decisions and consult with third parties to justify the use of medical techniques and billing. So we see that a decade ago, “physician” and “administrator” referred to two different people. Today, physicians realize that they have to embrace both titles, and their separate (and sometimes opposing) strategies (Lazarus, 1999; Lipson, 1997).
When we look at the importance of savings from computer adoptions and compare doctors who spend less than 10% of their time on the business aspects of their practice, we find that there is no correlation between the importance of various types of usage in the 1994 vs. 2003 (Table 9). In 1994, the three most important savings were improving insurance claims processing, increasing cash flow, and reducing administrative overhead. In 2003, access to patient hospital information, increasing cash flow, improving patient record keeping, and use of computer technology was the highest. Thus, the doctors in 2003 who spent little time on business stressed the patient information and record keeping. When we look at the doctors spending more than 10% of their time on business aspects in 1994 and 2003, we also find little correlation between the various types of usage. In fact, we find a somewhat negative correlation in what they found as the most important savings. In 1994, improving insurance claim processing, reducing administrative overhead, increasing cash flow, and financial performance controls were most important. In 2003, access to hospital patient hospital information and increasing cash flow were first and second in importance, and the use of computer technology came in third. These doctors in 2003 were also interested in reducing service bureau costs and image storing of patient records, which were not concerns for their counterparts in 1994 since image storage was not common in the early 1990s. Moreover, patient record keeping has grown tremendously in importance probably due to the new HIPPA legislation (Jonietz, 2003).
If we compare the two groups (under 10% and 10% plus) for 2003, there is a very high correlation (rho = 0.957 and p = 0.0000003, bottom of Table 9). Thus, for each year, separately, the time spent by doctors on business matters doesn’t seem related to the importance they accord to different business aspects. However, over the decade, different business aspects have become more important.
This last assessment seems to be the overarching theme of the data. The changes that have occurred in IT usage during the last decade have been in response to physicians trying to adapt to greater business demands on them. Physicians are becoming more aware of this with new programs in medical schools focusing on business. Texas Tech, for example has instituted a combined MD and MBA program. As physicians become busi-nesspeople as well as healers, they have had to embrace information technology in different ways. Apparently, the profession as a whole is adapting to this change. This is seen in Table 10, where the amount of business information provided by associations/professional meetings, professional journals and vendors has increased statistically significantly from 1994. On the other hand, the use of consultants has decreased significantly from 1994, where it was the number one source of information. This shows that physicians no longer have to pay experts to provide business information to them but are being supported by their profession.
conclusion
Since the 1970s, researchers and practitioners have attempted to apply IT to the practice of medicine in order to increase efficiencies and decrease costs; however, the work is far from over. Technological leaps such as the Internet and the World Wide Web have been born and begun to impact medicine (Prater & Roth, 2002). The medical industry still has no clear common goals for IT and very few universally accepted standards. More than 90% of the $30 billion in health transactions are carried on via phone, fax or paper (Shine, 1996). As an example, in only 22% of clinics in North America can a clinician call up a medical record, input information, and enter orders (2002). To quote Tommy Thompson, the former U.S. Secretary of Health and Human Services, “some grocery stores have better technology than our hospitals and clinics” (Turner, 2004). Other researchers have argued that we are beginning to see the first-level benefits of digitization such as increases in speed, control, accountability, and cost containment (Flower, 2003). This line of thinking is supported by the new federal 10-year initiative to “use Medicare as a vehicle for pilot programs ranging from handling prescriptions electronically to moving patient records online so that caregivers and patients can refer to them regardless of time or place” (Turner, 2004). Our paper shows that physicians have, in increasing numbers, embraced the use of IT in these areas. Thus, one of the benefits of this paper is to show the growth and trend lines of the various types of IT use by physicians for the past decade. We have also noted the move from primarily transactional uses of IT to informational and strategic uses.
Also, we have been able to look at the changes in infrastructure for group medical practices over the past decade. We have compared the different applications used over the decade by different-sized practices and by relatively new versus old practices. We have also measured the importance of different types of savings from computer adoptions and shown that, especially for small practices, the importance of these types of savings have changed significantly. Finally, we have shown the tremendous increase in time that physicians must devote to the business aspects of their practices.
We see a trend from the use of IT in transactional uses such as billing and insurance company dealings to more informational uses as patient record and contact with hospitals (Dutta & Heda, 2000). There has also been an increase in strategic emphasis of IT. This would include such uses as enhancing professional image use of expert systems and business planning, which have significantly increased in importance over the decade (Tables 3, 6). Moreover, we see big changes in infrastructure as discussed by Weill and Broadbent (1998). The number of practices with network-connected PC’s has doubled. Mainframe and mini-computers, which were used in 30.7% of practices in 1994, were used in only 12% of practices in 2003. Thus, the role of computers and the type of computer facilities employed in group medical practices have appreciably changed over the past decade.
One of the major problems of moving to IT systems is cost. A recent article in USA Today (Schmit, 2004) points out that wider use of computer software and hardware costs $10,000 to $20,000 per doctor and most of the benefits go to the insurers and hospital records. Doctor productivity drops 20% during the first 3 to 6 months after computer installation. Some possibilities for improvement are suggested by Bridges to Excellence, an employer coalition including GE and Ford Motor formed last year, which pays eligible doctors $50 per patient, per year, to use technology to improve healthcare.
While this paper provides a benchmark for the past decade’s use of IT by physicians, its best use is as a foundation for further research. First of all the consequences of the new HIPPA legislation may lead to more informational uses of IT in the physician’s offices (Jonietz, 2003). Second, while the overall use of IT in key clinical functions remains low (Networks, 2000, 2001), new technologies are being developed on an almost daily basis. Electronic patient records are being utilized more frequently (Hassey, Gerrett, & Wilson, 2001; Mondl, Szolonits, & Kohane, 2001). There is a need to determine the use of computerized physician order entry (CPOE) systems, disease registries, and pharmaceutical surveillance systems. Patient-centered Internet applications are expanding. An example is Pa-tientSite, which allows access to patient’s medical records (except for clinical notes). It also allows patients to schedule appointments online. Other systems tie in hospital-based IT with private practitioners. An example would be eICUs where hospital specialists can remotely track as many as 105 patients in intensive care. If problems arise, direct intervention can be signaled to the patient’s on-site medical staff, while the personal physician is being called (Turner, 2004). Another example is hospitals that provide online prescription renewal. These examples compose just a partial list. The key issue is that physicians are just beginning to utilize IT. In a recent study by BCG reported in the Wall Street Journal (Landro, 2002), only 42% of U.S. physicians are using electronic records or plan to do so in the near future. However, others have argued that, “at the end of the day, a physician’s value will depend on whether he or she can still connect to patients in this cyberspace odyssey” (Healy, 2004).
So will the majority of physicians end up going the same route as the early movers or will they demand new IT capabilities? Will these new applications of IT come from the U.S. or overseas? Given that IT is interconnecting the world and governments worldwide strive to provide excellent medical services while containing costs, there is a need to glean information from worldwide practices. As of now, we do not know, but we will be applying an expanded form of this study in Taiwan this summer. The benchmarking done in this study, combined with periodic research assessing changes will provide the tools that researchers and practitioners need to best anticipate and manage the changes in medical practices that the future holds.
