- Integrated Ambulatory Care Services in Oncology John P. Enterline, M.S., Farideh M. Majidi, M.S., Lisa Lattal Ogorzalek, M.H.A., J.D. Gloria J. Stuart, Jerome S. Rauch, M.L.S., Monica D. Fulton, R.N., M.B.A. David S. Ettinger, M.D. The Johns Hopkins Oncology Center, Baltimore, Maryland ABSTRACT In today's medical care environment of cost containment and restricted reimbursement, it is important to maximize the use of expensive facility and personnel resources. Concurrently, it is important to provide superior and timety patient services in order to remain competitive in an extremely flexible market. There are many areas in today's larger hospital environments where such ideals can be easily achieved. One of the more obvious areas is the automation of appointment and resource scheduling for ambulatory care services. This article focuses on maximizing the use of available physical and personnel resources in the ambulatory care setting of large and specialty hospitals. The Johns Hopkins Oncology Center's integrated outpatient scheduling and resource management systems are used as examples of what can be achieved. It is hoped that the experiences of the Oncology Center in developing these integrated systems will help others in similar efforts.
INTRODUCTION Sophisticated hospital scheduling and resource management systems for both inpatient and outpatient settings have lagged behind comparable systems in other industries. Examples of well defined scheduling and resource management systems include manufaurig, nsportation, banking, finance, insurance, airport facilites, and commcations. Over the past decade, these industries have made significant investments in computer-based automation in order to improve efficiency and remain competitive. During this period, hospital investments in computer automation have remain relatively low. Historically, the major focus in hospital infonnation systems has been on financial and billing systems rather than clical, scheduling, and resource management applications. Yet there is no industry in greater need of automaton than the hospital clinical-care environment. Particuariy in the over-crowded, and inefficient ambulatory care environments of many large and specialty hospitals, automated systems will become essential to remain cost effective. Many larger health c facilities view ambulatory care scheduling and management as functions which cab be performed manually. When automated systems exist for scheduling, they generally do not integrate with other clinical, inistrative, and charge capt funcons. 0195-4210/92/$5.00 01993 AMIA, Inc.
197
In the late 1970s, The Johns Hopkins Oncology Center recognized the need for a computer-based clinical decision-support system which structured information on patients as a care continuum rather than a series of unrlated event-specific data. This was due to both the relatively long periods of time that cancer patients were under complex medical care, and the enormous amounts of data which were generated by extremely ill patients with concomitant diseases. As a result, the Oncology Clinical Information System (OCIS) was developed to help with the complex task of decision making witiin the Center.' Eady in the OCIS development process, it became apparent that the integration of functions, such as ancillary services, iive utilities, research systems, and financial processes was a logical and natural extension to the basic patient-time oriented database structure. Due to the highly strumred pplications and database approach used in systems' development, adding new functions in a modula and maintainable manner was possible.2 Subsequently, a series of related support applications were integrated into the database and applications structure of OCIS. The focus of the present paper is on a varety of integated sub-systems which suppor the automation requiements of the ambulatory care component of The Johns Hoins Oncology Center. This support has become mcreasingly important as the poporon of primary care given in an outpatient setting has grown. The following system description should provide an example of how modem ambulatory sevices are a direct and logical extension of other sces provided at a hospitaL In the future medical care reimbursement environment, it will be critical to provide efficient, effective, and timely patient care through similar integrated systems in order to remain cost-competitive.
SYSTEM DESCRIPTION One of the many ancillary support applications developed around the OCIS strumcre is an OPD appointment scheduling, resource maagement, and charge-capure system. This OPD system is basically an operational decision-support system which uses data from a variety of clinical and ancillary systms in the Oncology Center. Concrenty, it feeds other systems with impant information on appointnents, tests, and procedures. Rarely does a patient in Oncology have an event scheduled that does not relate to a wide variety of other information. In general, each
scheduled event has a close relationship with other scheduled events, outcomes of clinical tests, clinical progress, available resources, clinical protocol requirements, and events scheduled for fuur visits. The operational flow and management of outpatient services within Oncology are defined and conducted through this system. Patients, care providers, and ancillary support personnel know what events to expect, and when to expect them. All procedure, charge, and clinical data are automatically collected in a complete, accurate, and timely manner and integrated into patient files. Additionally, this system provides a means of communication with other ancillary and clinical resources which are essential in the effective and efficient treatment of ambulatory patients. Such areas include the phannacy system, blood product systems, inpatient services, laboratory systems, research systems, and other clinic systems. Such scheduling, decision-support, and data distribution capabilities would not be possible in an information systems' environment that did not provide fully integrated access to all ancillary and clinical information. Thus, the root of the system's integration is its patient and time oriented database, and applications structure. The philosophy of the OCIS system has been described previously.3
Organizational Structure The ambulatory care component of the Oncology Center consists of four physically and logically unique clinics: Medical Oncology, Pediatric Oncology, Radiation Oncology, and a Medical Oncology followup/consultation clinic. The ability to communicate and share data in real time is essential in this distributed environment Patients fequently have appointments in multiple Oncology clinics during their course of treatment and follow-up at the Center. Each of these clinics has a core of scheduling requirements which are identical, combined with an assortment of requirements which are unique to the clinic. Most of the scheduling applications are driven by a common database engine. Additionally, these clinics have sub-clinics associated with them. These sub-clinics have additional requirements which are unique. Each clinic and associated sub-clinic also has a variety of care providers including physicians, nurses, and technicians. Providers can be members of multiple clinics. The scheduig flexibility for each group of providers is unique. Within groups, individual care providers have unique scheduling requixements. As a primay care physician can be included in more than one chnic, all scheduling data for a physician can be viewed across clinics by all regista during the appointment scheduling process. The primary Medical Oncology clinic is a reasonable representation of other clinics in Oncology. Therefore, the functions of this clinic are described below to provide an understanding of how
198
appointnent scheduling and the other computer-based support operates.
Ordering an Scheduling Systems Medical Oncology has a number of different scheduled event types. These include primary physician appointments, nursing appointments, procedure appointments, test ordering, and treatnent appointments. Sometimes these appointments must be linked with the schedules of other clinics, e.g., Radiation Oncology. In most instances, patients are scheduled for multiple events within a clinic, e.g., blood drawing, physician, nurse, chemotherapy treatment. These events must be coordinated with regard to sequence and necessary lag-times between events. Each of these scheduled event types is handled by Medical Oncology OPD registras through a system which is integrated with all other appointment and clinical systems. Registration and Appointment Scheduling. Medical Oncology has 45 full time registran and two new patient coordinators to handle approximately 28,000 patient visits a year. Again, each of these patient visits generally involves several appointments. Prior to an initial patient visit, essential information is collected and entered into the system. This includes demographic data, insurance data, and outside clinical infonnation. A complete update of insurance and demographic information is required by the system every thre months. Most patients treated in the Center start with an outside referal. A new patient is ocra WIO
"D. DUZ l m TXUXE
1/9/92
xTZcU AWCZZ&hR Apommx
SDLZC ana UW DcE
B
am :
NZLR
CUL
a~~~~~~~c
31w
.R.N: .A.R. . ZDWLLTZCU 3MZ .a .I.Oi: l
Figure 1. Main OPD Appointment Scheduling Menu. assigned a primary physician based on the patient's disease, stage of disease, potential clinical protocols, and the availabiity of new patient openings in a physician's schedule. Patients are also assigned a
primary nurse that is automatically linked with the patient's primary physician. All appointment scheduling is performed on-line using a variety of scheduling tools (Figure 1). These tools allow the registrar to view available patient, resource, and provider data in order to make a rational scheduling selection. In general, regiss will first schedule a primary physician appointnent for the The system will automatically link the patient. patient's existing schedule with his primary physician's schedule. If a nursing visit is scheduled, the patient's schedule is also linked with the nurse's schedule in a simlar manner. When the primary appointnent is selected, the registrar is prompted to schedule any nursing and other ancillary appointnents. These ancillary appointments can be for procedures, tests, radiology, scans, and treatment Depending on which ancillary appntments are selected, the times for these appointments are automatically selected either before or after the primary appointment. Necessary lag times between appointments are accounted for by the system. The regisas schedule the best available time slot based on full knowledge of available local resources and the The system patient's schedule in other clinics. provides guidelines and wamings for the registrar which assists in remembering the mles associated with different events. The system will not allow certain scheduling sequences or over-bookings for conflicts which are considered cxitical, e.g., scheduling a treatnent too close to an exam. Registars also have the ability to schedule appointments for multiple visits during one resation session. When necessary, primary and ancillary appointments can be modified or canceled. Block functions allow regisar to reschedule series of patient specific appointments during a single session.
Primary provider template clinic. available
Appointment Schedules. Each primary modify his or her baseline schedule according to the gudelines of a specific This lets registrars know when times are for on-line scheduling, and whether new patients can be accepted. Physicians and nurses can modify their schedules to block days and times where new and return patients can be seen. The physician can limit the number of new and return patients which can be scheduled in a set period of time. Tbe hours available for appointments, the tpes of appointments, and the time allotted for each appointment can also be modified. Schedules are set up to include both standing meeting and vacation times. can
Test and Procedure Ordering. Tests and procedures can be ordered at the time of appointnent scheduling. Regista are prompted to order the approriate tsts and procedures, depending on what appointments have been scheduled, and what clinical data are necessary based upon the Appointment Slips filled out by
physicians or nurses. Additional tests and procedures can also be docwnented duing the primary appointment on the Routing Fonn described below. Treatments and procedures which require special environments are generally ordered through the Chemotherapy and Treatment Facility schedulig system also described below. Group ordering functions allow logical groups of tests and procedures to be ordered through a single selection. Medications are ordered through the unit-dose OCIS pharmacy system which supports both the inpatient and outpatient populations.4 It is important to note that the schedules of tests and procedures default to times relative to the patent's primary appointment. For example, if blood work is ordered in conjunction with a primary appointment, the default time scheduled for the blood draw precedes the primary appointnent by 20 minutes. This 20 minute lag allows adequate time for the test results to be
C.atoufqi 3:00 AXto 1:45 me r 0a/i*/92 --_eg Ohaz_3d Ca3: Sm: VS: lO: 1O: la:g103: 10: lot: 10g: 1:110:1iX0:gi0: In: 1:i:gia:1ai:1 OOiS1gSOag0SsgOOallsgl3s l45a00sgi5z@ag5l0g45sg00sg3ia30ag45I caiui *g gIM gu gIn=If3ggIng Ila g=g=l 31 1t3 IXlal=g C2s/ Ing=g=mg. -i ...gI= gingmgIM uIM I
us/us DOW
u2 gus2 IgIngIMg IN2 1112 I11 1 lR32 1 In g 1gs in .g"w.2 1 11I I 1"i ///i|///aI/// IJ/J/I///gI/// I///gI//JI/g////g////g////g////g/-///g// Igu///Ig/ mu gus gmginggingmgingingmgmgg///g///,//////I "'g1gu gu ggingmgingmg/////ug/gu/g//////g/ugy/g gus g=gm=1 ///I///J///J///g///g///I OMJIM
5634
34 3
C4/s2 cs/s csius
si/u3
37/3
as/u S/"r (anes
giginginmgi
"'gli3
134 gigigigigm///g///g////
134 gininggmg'g3
gnmgm
l///t///i
11 L11/ *.J*J****.g*wg4gI =34 1=1=11311" mg=J=Jmgu4 gmJ=ggX4lI //////gI//////I///I///g ////// /// /// ///gI//g/I//g/I/
///s///
Figure 2. Treatment Scheduling Grid (Screen). available to
physician for assessment during the If chemotherapy is ordered on a day when a primary appointment is being scheduled, the schedule for the chemotherapy is placed after the primary appointment. The lag time between the primary appointment and the chemotherapy appointment depends on other events which are scheduled, and the estimated time the phannacy needs for preparation of the chemotherapy formulation. the
prnmay visit.
Treatment Scheduling. The system in Medical Oncology which handles chemotherapy and other tament scheduling is more complex than the systems which schedule a provider appointment, procedures, tests, and medications. However, this system is fully integrated with all other scheduling systems and elies on all information during the scheduling process. The registrar aatually views all scheduling and resource management subsystems as a single unified system. There are over 125 different types of chemotherapy and procedures routinely given in the Center's
199
Each of these ambulatory treatment facilities. therapies requires varying amounts of resource time, nursing start-up time, and nursing monitoring time. Additionally, approximately 30% of the patients treated There are 6 require more than one procedure. treatment chairs and 14 treatment beds staffed by 5-7 nurses, depending upon the patient load and the types of treatments scheduled. Each nurse must cover 3-5 Each treatnent resources and associated patients. nurse and each physical resource can only handle a limited number of appointments, depending on the A sample scheduling grid is types of treatments. presented in Figure 2. This provides a visual aid in selecting an appropriate time for a specific therapy. Finding an optimal fit for a chemotherapy appointment requires an understanding of multiple This subsystem provides all dynamic variables. necessary data for the optimal selection of schedules across resources. Approximately half of the time, chemotherapy treatnent appointnents are preceded by a primary physician appointnent and blood tests. As previously mentioned, scheduled patient appointments, tests and procedures are appropriately sequenced around the primary provider appointnent. To furither complicate matters, chemotherapy can be a form of preparation for radiation therapy. This requires close coordination with the Radiation Oncology scheduling systems. Radiation Oncology systems are comparable to the scheduling and support systems available to Medical Oncology.
Patient-Provider-Facility Relations. Each clinic handles the relationship between providers, treatment In Medical resources, and patients differently. Oncology, each primary physician is assigned a primary nurse. Each clinic patient is assigned a primary physician and the associated primary nurse. These links are maintaied on the scheduling system. Relations are also necessary between clinic-specific scheduling systems. A good example of this is the scheduling link needed between the chemotherapy system and the radiation system. Tissues are sensitized by chemotherapy to allow radiation to be effective. Tight "windows of sensitivity" require that the chemotherapy be provided dixectly prior to radiation therapy. This requires very tight scheduling of: (1) all necessary tests and procedures, (2) required primary provider appointnents, (3) orders to the phannacy for the required chemotherapy, (4) administration of chemotherapy, and (5) radiation
therapy. Special Features and Attnrbutes There are a wide variety of extensions to the basic scheduling and ordering systems. These include extensive automatically generated forms and reports, charge-captre capabilities, and resource loadmonitoring/forecasting procedures. A few of these extensions are described below.
200
=-: s-
Forms and Reports. As mentioned, the scheduling system provides the foundation for general operations in the OPD. Based on the schedule of patient appointments, a variety of hard-copy reports and documents are automatically generated by OCaS. These include (1) clinical data flows, (2) routing forms, (3) various kept appointment reports, (4) several types of patient chart pull lists, (5) physician and ancillary staff schedules, (6) patient appointment lists, and (7) a wide variety of reports to assist OPD staff and ancillary resources. Daily, over 50 unique types of hard-copy reports in multiple copies are provided for the Medical Oncology clinic. A wide varety of other hard-copy reports are ordered as necessary. One of the most important decision-support tools provided by OCIS is automated patient data flows .~ma3xmrs 03091001Y
3R283C31 N0: 7777776 03t11va US, D: 02/28/52
3ccmmpzjL MOW228m oa
0
s
-91 C! O* s
-: :--s--5 4: 7:5 0:3 51----4-:--5003 92 2i2: :21in2 :2n2:2U2:2i2M2U2:2 *s6IX sDA 4&sDA 47tDAY 4$tMZ 4P:tl= Us= bit
s 2
"131 3m
sOll :031
s-
40:s0
15:03
41:M6 34s:3
42s:02 4a1:0 17:DW1 36303
1 C: 8W4R 37.4: 17.4: 27.:o 17.0: IN 1WS: 11d/SO: 130/SO: 124/"5 140/S0: 053 1U1m 94: : 100: 10: *=s 0 : 10 IN= 20: I: 20: 20: w 3SM 70.2: 70.5: 70.2: 40.4: S: :.0 NM MR 1.57: 1.67: 1.80: 1.65:
:1
R:
-3 :r.Ot
a
#/Co s C/0 0/DL0 : :
:030 :2W 0 :31
WZ
i
Io&"
mm
0U Sam x to= C. s:W 002
MU/I:
*:
073*
3a.50* 10.2*
096 1.41* 10.0*
110* 3.15' 10.1'
354' 2.49' 10.1'
29.5'
0.4'
29.4'
:2 7:
:
137:
116:
1.9:
102:
UOJL:
27:
103: 31'
a
730 72:
720: -72 I
400: :
100l:
000: M000:
1000:
us u
:t
435slA 44: 400t3 41:
17.0: 37.4: 122/S0: 130/0:
0:t 20:
De: 20:
5s.5:
4o.5: 1.06:
1.65: s
s
70.0:
5.62:
:5s.82::
4.1:
:MWU-S 3
D3103
:
3/: WUQ/1.:
:3010.aY3 1W
:31300010
s
/60: 26000* 45000* 14000' .-
s
.09:
44aW 35:031
134: 4.1: 104:
20.t9 22000' 13:
136:
110' 2.16' 9.4'
100' 1.20:
6.0w
26.1' 27.7* 24.1* 30000' 23000' 25000' 110: 1:
1s
116:
4.2:
4.1:
101: 22'
104:
7n: 72:
720:
600:
22'
220' 1.27* 9.40
117:
140:
4.4:
4.9:
27:
102: 321
101: 25s
72:
720: 72:
720: 72:
600: 1000:
600:
0o0:
1000:
1000:
a
Figure 3. Sample Clinical Flow Report (Minimal). (Figure 3). These flows combine inpatient, outpatient, and extemal data into a single, user-defined set of time-oriented data. The flows contain a wide variety of information in columns by date. This information includes all test results, vital signs, counts, treatment data, micro-bacteriology results, antibiotics, and medication infomnation. While the example flow sheet contains only a few variables, a full flow can contain hundreds of data elements which are collected daily. There are over 100 flow fonnats to meet unique clinical and research requirements. A logistically complex system has been developed which ensures that the most recent data are included in flows, and that these flows are available in the charts on the medical records pull lists. These flows are incorporated into the patient charts and provide a "story-book" type history of all patient data ever
entered into OCIS. It should be noted that patient data in OCIS are never archived or deleted The Routing Form is another important document generated by the scheduling system. Routing fonns are generated the night prior to appointnents and include all scheduled appointment, test, and procedure infornation as well as all necessary patient information. The routing form includes an extensive check list of additional procedures which that patient may require during a visit The Routing Forms follow patients through the course of all scheduled and
unscheduled events during a day. Tests, procedures, appointnents, and therapies which are "kept" are so indicated on the Routing Form. Any additonal events which occur are checked off or noted on the Routing Forms. The routing form provides a cross-check with what was scheduled for a patient visit during a post-visit verification process. It provides information on what additional services were provided to the patient on that visit. At the end of each day, Routing Forms are collected and matched against the automated "kept" appointment lists. These are generated from patient check-in data. Missing forms from kept appoinanents can immediately be identified and located within the department If the Routing Fonns can not be located, the days events can be reconstructed from OPD staff and OCIS records. Interestingly, a paper-less Routing Form will have been piloted in the Radiation Oncology clinic by the tie this paper is in print. In this system, all patient clinical care events will be recorded by the provider at the point of contact If this system is successful, it will be implemented in other clinics across the Oncology Center.
Charge Capture System. The day following an appointment the Routing Form is used to ensure that all patient activities are capued by the OCIS system. This is the means of charge capure used by OC[S. Only events which were not scheduled need to be entered. If a scheduled event did not take place, it is removed from OCIS. Billing information is automatically sent to the both the professional fees and hospital billing systems. All necessary procedure codes are built into data dictionaries within OCIS for proper coding. This system provides both on-line and retrospecive views of services which assists in quality control and patient billing problems. The benefit of the linked scheduing, routing, and charge capture systems is a dramatic increase in outpatient revenue through more complete, accurate and timely charges. e function Planning. Another important a of the appointment scheduling system is to provide historical data for forecasting and plig purposes. Numerous reports are generated for daily monitoring and short term st planning purposes. Additional reports are routinely generated for longer-term
forecasting. Ad hoc queries are frequently requested for special purpose management studies.
Systems Integration The appointnent scheduling and resource management subsystens are integrated functions of OCIS. They depends on data generated through other o00s systems, and provides essential data back to these systems. In most instances, a patient is seen in the OPD prior to being admited as an inpatient Thus, there is ongoing need for tight communication between the OPD and the inpatient units. As all clinical data are integrated, all treatment, demographic, referral, vital signs and physical data entered in OCIS are readily available to both the inpatient and outpatient staff. When patients are converted from an outpatient to an inpatient, or visa versa, there is no need to collect a variety of information on that patient Thus, all Center personnel are automatically provided with timely and complete patient information. CONCLUSIONS Scheduling and resource management systems in an ambulatory care setting should not be viewed as a function which is independent from other clinical care activities. In order to opmize both personnel and physical resources, information not historically associated with single appointment scheduling is required. This is particularly important in environments where proper ambulatory care involves sequentially scheduling groups of related events, procedure, and resources over time. Additionally, moving patients through the medical system in a timely manner will be increasingly important from the patients' perspective, and to remain cost-competitive in an increasingly competitive market. The proper automation of ambulatory care scheduling and resource management will become critical in the futr. REFERENCES 1.A Clinical Information System for Oncology. Enterine, JP., Lenhard, RE., Blum, B.L Eds, Springer-Verlag, N.Y., N.Y., 1989.
2.TEDIUM and the Software Process. Blum, B.I., The MIT Press, Cambridge, Massachusetts, 1990. 3. Enterline, J.P., Lenhard, RE., Blum, BI . The Oncology Clinical lnformation Systen. In A Clinical Information System for Oncologv. Entere J.P., Lenhard, RE., Blum, B.I. Eds, Springer-Verlag, N.Y., N.Y., 1989. pp. 1-21. 4. Harwood, PM., Causey, J.P., Goldberger, S. Pharmacy System. In A Clinical Information Syste for Oncoloy. Eneine, JP., Lenhard, RE., Blum, BI. Eds, Springer-Verlag, N.Y., N.Y., 1989. pp. 139-160.
201
INTRODUCTION Sophisticated hospital scheduling and resource management systems for both inpatient and outpatient settings have lagged behind comparable systems in other industries. Examples of well defined scheduling and resource management systems include manufaurig, nsportation, banking, finance, insurance, airport facilites, and commcations. Over the past decade, these industries have made significant investments in computer-based automation in order to improve efficiency and remain competitive. During this period, hospital investments in computer automation have remain relatively low. Historically, the major focus in hospital infonnation systems has been on financial and billing systems rather than clical, scheduling, and resource management applications. Yet there is no industry in greater need of automaton than the hospital clinical-care environment. Particuariy in the over-crowded, and inefficient ambulatory care environments of many large and specialty hospitals, automated systems will become essential to remain cost effective. Many larger health c facilities view ambulatory care scheduling and management as functions which cab be performed manually. When automated systems exist for scheduling, they generally do not integrate with other clinical, inistrative, and charge capt funcons. 0195-4210/92/$5.00 01993 AMIA, Inc.
197
In the late 1970s, The Johns Hopkins Oncology Center recognized the need for a computer-based clinical decision-support system which structured information on patients as a care continuum rather than a series of unrlated event-specific data. This was due to both the relatively long periods of time that cancer patients were under complex medical care, and the enormous amounts of data which were generated by extremely ill patients with concomitant diseases. As a result, the Oncology Clinical Information System (OCIS) was developed to help with the complex task of decision making witiin the Center.' Eady in the OCIS development process, it became apparent that the integration of functions, such as ancillary services, iive utilities, research systems, and financial processes was a logical and natural extension to the basic patient-time oriented database structure. Due to the highly strumred pplications and database approach used in systems' development, adding new functions in a modula and maintainable manner was possible.2 Subsequently, a series of related support applications were integrated into the database and applications structure of OCIS. The focus of the present paper is on a varety of integated sub-systems which suppor the automation requiements of the ambulatory care component of The Johns Hoins Oncology Center. This support has become mcreasingly important as the poporon of primary care given in an outpatient setting has grown. The following system description should provide an example of how modem ambulatory sevices are a direct and logical extension of other sces provided at a hospitaL In the future medical care reimbursement environment, it will be critical to provide efficient, effective, and timely patient care through similar integrated systems in order to remain cost-competitive.
SYSTEM DESCRIPTION One of the many ancillary support applications developed around the OCIS strumcre is an OPD appointment scheduling, resource maagement, and charge-capure system. This OPD system is basically an operational decision-support system which uses data from a variety of clinical and ancillary systms in the Oncology Center. Concrenty, it feeds other systems with impant information on appointnents, tests, and procedures. Rarely does a patient in Oncology have an event scheduled that does not relate to a wide variety of other information. In general, each
scheduled event has a close relationship with other scheduled events, outcomes of clinical tests, clinical progress, available resources, clinical protocol requirements, and events scheduled for fuur visits. The operational flow and management of outpatient services within Oncology are defined and conducted through this system. Patients, care providers, and ancillary support personnel know what events to expect, and when to expect them. All procedure, charge, and clinical data are automatically collected in a complete, accurate, and timely manner and integrated into patient files. Additionally, this system provides a means of communication with other ancillary and clinical resources which are essential in the effective and efficient treatment of ambulatory patients. Such areas include the phannacy system, blood product systems, inpatient services, laboratory systems, research systems, and other clinic systems. Such scheduling, decision-support, and data distribution capabilities would not be possible in an information systems' environment that did not provide fully integrated access to all ancillary and clinical information. Thus, the root of the system's integration is its patient and time oriented database, and applications structure. The philosophy of the OCIS system has been described previously.3
Organizational Structure The ambulatory care component of the Oncology Center consists of four physically and logically unique clinics: Medical Oncology, Pediatric Oncology, Radiation Oncology, and a Medical Oncology followup/consultation clinic. The ability to communicate and share data in real time is essential in this distributed environment Patients fequently have appointments in multiple Oncology clinics during their course of treatment and follow-up at the Center. Each of these clinics has a core of scheduling requirements which are identical, combined with an assortment of requirements which are unique to the clinic. Most of the scheduling applications are driven by a common database engine. Additionally, these clinics have sub-clinics associated with them. These sub-clinics have additional requirements which are unique. Each clinic and associated sub-clinic also has a variety of care providers including physicians, nurses, and technicians. Providers can be members of multiple clinics. The scheduig flexibility for each group of providers is unique. Within groups, individual care providers have unique scheduling requixements. As a primay care physician can be included in more than one chnic, all scheduling data for a physician can be viewed across clinics by all regista during the appointment scheduling process. The primary Medical Oncology clinic is a reasonable representation of other clinics in Oncology. Therefore, the functions of this clinic are described below to provide an understanding of how
198
appointnent scheduling and the other computer-based support operates.
Ordering an Scheduling Systems Medical Oncology has a number of different scheduled event types. These include primary physician appointments, nursing appointments, procedure appointments, test ordering, and treatnent appointments. Sometimes these appointments must be linked with the schedules of other clinics, e.g., Radiation Oncology. In most instances, patients are scheduled for multiple events within a clinic, e.g., blood drawing, physician, nurse, chemotherapy treatment. These events must be coordinated with regard to sequence and necessary lag-times between events. Each of these scheduled event types is handled by Medical Oncology OPD registras through a system which is integrated with all other appointment and clinical systems. Registration and Appointment Scheduling. Medical Oncology has 45 full time registran and two new patient coordinators to handle approximately 28,000 patient visits a year. Again, each of these patient visits generally involves several appointments. Prior to an initial patient visit, essential information is collected and entered into the system. This includes demographic data, insurance data, and outside clinical infonnation. A complete update of insurance and demographic information is required by the system every thre months. Most patients treated in the Center start with an outside referal. A new patient is ocra WIO
"D. DUZ l m TXUXE
1/9/92
xTZcU AWCZZ&hR Apommx
SDLZC ana UW DcE
B
am :
NZLR
CUL
a~~~~~~~c
31w
.R.N: .A.R. . ZDWLLTZCU 3MZ .a .I.Oi: l
Figure 1. Main OPD Appointment Scheduling Menu. assigned a primary physician based on the patient's disease, stage of disease, potential clinical protocols, and the availabiity of new patient openings in a physician's schedule. Patients are also assigned a
primary nurse that is automatically linked with the patient's primary physician. All appointment scheduling is performed on-line using a variety of scheduling tools (Figure 1). These tools allow the registrar to view available patient, resource, and provider data in order to make a rational scheduling selection. In general, regiss will first schedule a primary physician appointnent for the The system will automatically link the patient. patient's existing schedule with his primary physician's schedule. If a nursing visit is scheduled, the patient's schedule is also linked with the nurse's schedule in a simlar manner. When the primary appointnent is selected, the registrar is prompted to schedule any nursing and other ancillary appointnents. These ancillary appointments can be for procedures, tests, radiology, scans, and treatment Depending on which ancillary appntments are selected, the times for these appointments are automatically selected either before or after the primary appointment. Necessary lag times between appointments are accounted for by the system. The regisas schedule the best available time slot based on full knowledge of available local resources and the The system patient's schedule in other clinics. provides guidelines and wamings for the registrar which assists in remembering the mles associated with different events. The system will not allow certain scheduling sequences or over-bookings for conflicts which are considered cxitical, e.g., scheduling a treatnent too close to an exam. Registars also have the ability to schedule appointments for multiple visits during one resation session. When necessary, primary and ancillary appointments can be modified or canceled. Block functions allow regisar to reschedule series of patient specific appointments during a single session.
Primary provider template clinic. available
Appointment Schedules. Each primary modify his or her baseline schedule according to the gudelines of a specific This lets registrars know when times are for on-line scheduling, and whether new patients can be accepted. Physicians and nurses can modify their schedules to block days and times where new and return patients can be seen. The physician can limit the number of new and return patients which can be scheduled in a set period of time. Tbe hours available for appointments, the tpes of appointments, and the time allotted for each appointment can also be modified. Schedules are set up to include both standing meeting and vacation times. can
Test and Procedure Ordering. Tests and procedures can be ordered at the time of appointnent scheduling. Regista are prompted to order the approriate tsts and procedures, depending on what appointments have been scheduled, and what clinical data are necessary based upon the Appointment Slips filled out by
physicians or nurses. Additional tests and procedures can also be docwnented duing the primary appointment on the Routing Fonn described below. Treatments and procedures which require special environments are generally ordered through the Chemotherapy and Treatment Facility schedulig system also described below. Group ordering functions allow logical groups of tests and procedures to be ordered through a single selection. Medications are ordered through the unit-dose OCIS pharmacy system which supports both the inpatient and outpatient populations.4 It is important to note that the schedules of tests and procedures default to times relative to the patent's primary appointment. For example, if blood work is ordered in conjunction with a primary appointment, the default time scheduled for the blood draw precedes the primary appointnent by 20 minutes. This 20 minute lag allows adequate time for the test results to be
C.atoufqi 3:00 AXto 1:45 me r 0a/i*/92 --_eg Ohaz_3d Ca3: Sm: VS: lO: 1O: la:g103: 10: lot: 10g: 1:110:1iX0:gi0: In: 1:i:gia:1ai:1 OOiS1gSOag0SsgOOallsgl3s l45a00sgi5z@ag5l0g45sg00sg3ia30ag45I caiui *g gIM gu gIn=If3ggIng Ila g=g=l 31 1t3 IXlal=g C2s/ Ing=g=mg. -i ...gI= gingmgIM uIM I
us/us DOW
u2 gus2 IgIngIMg IN2 1112 I11 1 lR32 1 In g 1gs in .g"w.2 1 11I I 1"i ///i|///aI/// IJ/J/I///gI/// I///gI//JI/g////g////g////g////g/-///g// Igu///Ig/ mu gus gmginggingmgingingmgmgg///g///,//////I "'g1gu gu ggingmgingmg/////ug/gu/g//////g/ugy/g gus g=gm=1 ///I///J///J///g///g///I OMJIM
5634
34 3
C4/s2 cs/s csius
si/u3
37/3
as/u S/"r (anes
giginginmgi
"'gli3
134 gigigigigm///g///g////
134 gininggmg'g3
gnmgm
l///t///i
11 L11/ *.J*J****.g*wg4gI =34 1=1=11311" mg=J=Jmgu4 gmJ=ggX4lI //////gI//////I///I///g ////// /// /// ///gI//g/I//g/I/
///s///
Figure 2. Treatment Scheduling Grid (Screen). available to
physician for assessment during the If chemotherapy is ordered on a day when a primary appointment is being scheduled, the schedule for the chemotherapy is placed after the primary appointment. The lag time between the primary appointment and the chemotherapy appointment depends on other events which are scheduled, and the estimated time the phannacy needs for preparation of the chemotherapy formulation. the
prnmay visit.
Treatment Scheduling. The system in Medical Oncology which handles chemotherapy and other tament scheduling is more complex than the systems which schedule a provider appointment, procedures, tests, and medications. However, this system is fully integrated with all other scheduling systems and elies on all information during the scheduling process. The registrar aatually views all scheduling and resource management subsystems as a single unified system. There are over 125 different types of chemotherapy and procedures routinely given in the Center's
199
Each of these ambulatory treatment facilities. therapies requires varying amounts of resource time, nursing start-up time, and nursing monitoring time. Additionally, approximately 30% of the patients treated There are 6 require more than one procedure. treatment chairs and 14 treatment beds staffed by 5-7 nurses, depending upon the patient load and the types of treatments scheduled. Each nurse must cover 3-5 Each treatnent resources and associated patients. nurse and each physical resource can only handle a limited number of appointments, depending on the A sample scheduling grid is types of treatments. presented in Figure 2. This provides a visual aid in selecting an appropriate time for a specific therapy. Finding an optimal fit for a chemotherapy appointment requires an understanding of multiple This subsystem provides all dynamic variables. necessary data for the optimal selection of schedules across resources. Approximately half of the time, chemotherapy treatnent appointnents are preceded by a primary physician appointnent and blood tests. As previously mentioned, scheduled patient appointments, tests and procedures are appropriately sequenced around the primary provider appointnent. To furither complicate matters, chemotherapy can be a form of preparation for radiation therapy. This requires close coordination with the Radiation Oncology scheduling systems. Radiation Oncology systems are comparable to the scheduling and support systems available to Medical Oncology.
Patient-Provider-Facility Relations. Each clinic handles the relationship between providers, treatment In Medical resources, and patients differently. Oncology, each primary physician is assigned a primary nurse. Each clinic patient is assigned a primary physician and the associated primary nurse. These links are maintaied on the scheduling system. Relations are also necessary between clinic-specific scheduling systems. A good example of this is the scheduling link needed between the chemotherapy system and the radiation system. Tissues are sensitized by chemotherapy to allow radiation to be effective. Tight "windows of sensitivity" require that the chemotherapy be provided dixectly prior to radiation therapy. This requires very tight scheduling of: (1) all necessary tests and procedures, (2) required primary provider appointnents, (3) orders to the phannacy for the required chemotherapy, (4) administration of chemotherapy, and (5) radiation
therapy. Special Features and Attnrbutes There are a wide variety of extensions to the basic scheduling and ordering systems. These include extensive automatically generated forms and reports, charge-captre capabilities, and resource loadmonitoring/forecasting procedures. A few of these extensions are described below.
200
=-: s-
Forms and Reports. As mentioned, the scheduling system provides the foundation for general operations in the OPD. Based on the schedule of patient appointments, a variety of hard-copy reports and documents are automatically generated by OCaS. These include (1) clinical data flows, (2) routing forms, (3) various kept appointment reports, (4) several types of patient chart pull lists, (5) physician and ancillary staff schedules, (6) patient appointment lists, and (7) a wide variety of reports to assist OPD staff and ancillary resources. Daily, over 50 unique types of hard-copy reports in multiple copies are provided for the Medical Oncology clinic. A wide varety of other hard-copy reports are ordered as necessary. One of the most important decision-support tools provided by OCIS is automated patient data flows .~ma3xmrs 03091001Y
3R283C31 N0: 7777776 03t11va US, D: 02/28/52
3ccmmpzjL MOW228m oa
0
s
-91 C! O* s
-: :--s--5 4: 7:5 0:3 51----4-:--5003 92 2i2: :21in2 :2n2:2U2:2i2M2U2:2 *s6IX sDA 4&sDA 47tDAY 4$tMZ 4P:tl= Us= bit
s 2
"131 3m
sOll :031
s-
40:s0
15:03
41:M6 34s:3
42s:02 4a1:0 17:DW1 36303
1 C: 8W4R 37.4: 17.4: 27.:o 17.0: IN 1WS: 11d/SO: 130/SO: 124/"5 140/S0: 053 1U1m 94: : 100: 10: *=s 0 : 10 IN= 20: I: 20: 20: w 3SM 70.2: 70.5: 70.2: 40.4: S: :.0 NM MR 1.57: 1.67: 1.80: 1.65:
:1
R:
-3 :r.Ot
a
#/Co s C/0 0/DL0 : :
:030 :2W 0 :31
WZ
i
Io&"
mm
0U Sam x to= C. s:W 002
MU/I:
*:
073*
3a.50* 10.2*
096 1.41* 10.0*
110* 3.15' 10.1'
354' 2.49' 10.1'
29.5'
0.4'
29.4'
:2 7:
:
137:
116:
1.9:
102:
UOJL:
27:
103: 31'
a
730 72:
720: -72 I
400: :
100l:
000: M000:
1000:
us u
:t
435slA 44: 400t3 41:
17.0: 37.4: 122/S0: 130/0:
0:t 20:
De: 20:
5s.5:
4o.5: 1.06:
1.65: s
s
70.0:
5.62:
:5s.82::
4.1:
:MWU-S 3
D3103
:
3/: WUQ/1.:
:3010.aY3 1W
:31300010
s
/60: 26000* 45000* 14000' .-
s
.09:
44aW 35:031
134: 4.1: 104:
20.t9 22000' 13:
136:
110' 2.16' 9.4'
100' 1.20:
6.0w
26.1' 27.7* 24.1* 30000' 23000' 25000' 110: 1:
1s
116:
4.2:
4.1:
101: 22'
104:
7n: 72:
720:
600:
22'
220' 1.27* 9.40
117:
140:
4.4:
4.9:
27:
102: 321
101: 25s
72:
720: 72:
720: 72:
600: 1000:
600:
0o0:
1000:
1000:
a
Figure 3. Sample Clinical Flow Report (Minimal). (Figure 3). These flows combine inpatient, outpatient, and extemal data into a single, user-defined set of time-oriented data. The flows contain a wide variety of information in columns by date. This information includes all test results, vital signs, counts, treatment data, micro-bacteriology results, antibiotics, and medication infomnation. While the example flow sheet contains only a few variables, a full flow can contain hundreds of data elements which are collected daily. There are over 100 flow fonnats to meet unique clinical and research requirements. A logistically complex system has been developed which ensures that the most recent data are included in flows, and that these flows are available in the charts on the medical records pull lists. These flows are incorporated into the patient charts and provide a "story-book" type history of all patient data ever
entered into OCIS. It should be noted that patient data in OCIS are never archived or deleted The Routing Form is another important document generated by the scheduling system. Routing fonns are generated the night prior to appointnents and include all scheduled appointment, test, and procedure infornation as well as all necessary patient information. The routing form includes an extensive check list of additional procedures which that patient may require during a visit The Routing Forms follow patients through the course of all scheduled and
unscheduled events during a day. Tests, procedures, appointnents, and therapies which are "kept" are so indicated on the Routing Form. Any additonal events which occur are checked off or noted on the Routing Forms. The routing form provides a cross-check with what was scheduled for a patient visit during a post-visit verification process. It provides information on what additional services were provided to the patient on that visit. At the end of each day, Routing Forms are collected and matched against the automated "kept" appointment lists. These are generated from patient check-in data. Missing forms from kept appoinanents can immediately be identified and located within the department If the Routing Fonns can not be located, the days events can be reconstructed from OPD staff and OCIS records. Interestingly, a paper-less Routing Form will have been piloted in the Radiation Oncology clinic by the tie this paper is in print. In this system, all patient clinical care events will be recorded by the provider at the point of contact If this system is successful, it will be implemented in other clinics across the Oncology Center.
Charge Capture System. The day following an appointment the Routing Form is used to ensure that all patient activities are capued by the OCIS system. This is the means of charge capure used by OC[S. Only events which were not scheduled need to be entered. If a scheduled event did not take place, it is removed from OCIS. Billing information is automatically sent to the both the professional fees and hospital billing systems. All necessary procedure codes are built into data dictionaries within OCIS for proper coding. This system provides both on-line and retrospecive views of services which assists in quality control and patient billing problems. The benefit of the linked scheduing, routing, and charge capture systems is a dramatic increase in outpatient revenue through more complete, accurate and timely charges. e function Planning. Another important a of the appointment scheduling system is to provide historical data for forecasting and plig purposes. Numerous reports are generated for daily monitoring and short term st planning purposes. Additional reports are routinely generated for longer-term
forecasting. Ad hoc queries are frequently requested for special purpose management studies.
Systems Integration The appointnent scheduling and resource management subsystens are integrated functions of OCIS. They depends on data generated through other o00s systems, and provides essential data back to these systems. In most instances, a patient is seen in the OPD prior to being admited as an inpatient Thus, there is ongoing need for tight communication between the OPD and the inpatient units. As all clinical data are integrated, all treatment, demographic, referral, vital signs and physical data entered in OCIS are readily available to both the inpatient and outpatient staff. When patients are converted from an outpatient to an inpatient, or visa versa, there is no need to collect a variety of information on that patient Thus, all Center personnel are automatically provided with timely and complete patient information. CONCLUSIONS Scheduling and resource management systems in an ambulatory care setting should not be viewed as a function which is independent from other clinical care activities. In order to opmize both personnel and physical resources, information not historically associated with single appointment scheduling is required. This is particularly important in environments where proper ambulatory care involves sequentially scheduling groups of related events, procedure, and resources over time. Additionally, moving patients through the medical system in a timely manner will be increasingly important from the patients' perspective, and to remain cost-competitive in an increasingly competitive market. The proper automation of ambulatory care scheduling and resource management will become critical in the futr. REFERENCES 1.A Clinical Information System for Oncology. Enterine, JP., Lenhard, RE., Blum, B.L Eds, Springer-Verlag, N.Y., N.Y., 1989.
2.TEDIUM and the Software Process. Blum, B.I., The MIT Press, Cambridge, Massachusetts, 1990. 3. Enterline, J.P., Lenhard, RE., Blum, BI . The Oncology Clinical lnformation Systen. In A Clinical Information System for Oncologv. Entere J.P., Lenhard, RE., Blum, B.I. Eds, Springer-Verlag, N.Y., N.Y., 1989. pp. 1-21. 4. Harwood, PM., Causey, J.P., Goldberger, S. Pharmacy System. In A Clinical Information Syste for Oncoloy. Eneine, JP., Lenhard, RE., Blum, BI. Eds, Springer-Verlag, N.Y., N.Y., 1989. pp. 139-160.
201
