Birth Center Outcomes Reported Through Automated Technology

J

0GNN

CLINICALISSUES

Birth Certter Outcomes Reported Through Automated Technology Leissa Roberts, RNC, CNM, MS, Katherine Sward, RN, M S

=

Clinical Outcomes Research

Accepted: June 2000

The need to measure clinical outcomes has long been recognized by health care payers, institutions, and providers. Clinical data must be provided to state agencies and regulatory bodies such as the Joint Commission on Accreditation of Healthcare Organizations (Johnson, 1994; Zielstorff, 1995). The Health Care Financing Administration and the Health Plan Employer Data and Information Set publish guidelines for reporting key quality indicators. Practitioners must also provide outcomes data to consumers, who are demanding increased accountability from their health care professionals. The National Birth Center Study (NBCS) was a large, multicenter study of clinical outcomes in birth centers (Rooks et al., 1989). Although more than 10 years old, this study of nearly 12,000 women in 84 birth centers was one of the largest published birth center studies. The study described a variety of clinical outcomes and concluded that birth centers were a safe alternative birthing site. The NBCS reported maternal factors such as age, ethnicity, and parity and infant factors such as Apgar score, birth weight, and gestational age. Other outcomes reported were mortality rates, the rate of transfer to the hospital, and maternal and infant complications. Two years later, Albers and Katz (1991) reviewed the literature on birth settings and concluded that nontraditional settings such as birth centers provide a safe, cost-effective, alternative birthing site. Fullerton et al. (1997) compared transfer rates from freestanding birth centers to those reported in the NBCS. Those authors noted that factors unique to each practice site could influence clini-

Birthcare Healthcare is a multidisciplinary health care clinic and freestanding birth center in Salt Lake City, Utah. Highlights of a woman's labor assessment and birth outcomes are summarized in a paper birth log. The authors developed a relational database that facilitates efficient outcomes reporting. Examples of outcomes from this database are reported. Objectives: This study describes clinical outcomes from a freestanding birth center and the automated birth log database used to manage outcomes reporting. Design: Retrospective data obtained primarily from a paper birth log and secondarily from the paper chart were entered into and extracted from an electronic database. Participants: 23 1 women who were evaluated in 269 encounters for delivery at the birth center between July 1 , 1997, and June 30, 1999. Main Clinical Outcome Measures: The number of deliveries at the birth center; the percentage of intrapartum, postpartum, and newborn transfers to the inpatient hospital setting; cause-specific transfer rates; method of transfer; medication in labor; low and high birth weights; percentage of low Apgar scores; and average times from admission to birth and birth to discharge. Conclusion: Birth center clinical outcomes can be efficiently reported using an automated birth log. Clinical outcomes reported in this study are consistent with finding of previous research that demonstrates that birth centers are a safe alternative site for delivery. 10GNN, 30, 110-1 20; 2001. Keywords: Birth center-Birth log-CNM outcomes- Database

110 JOG"

Volume 30, Number 1

cal outcomes and suggested that the NBCS and other published reports can serve as points of reference but might not be appropriate benchmarks for evaluating the clinical practices of a particular site. The automated birth log for Birthcare Healthcare (BCHC) was developed to facilitate internal benchmarking data reports. Although practice sites have unique characteristics and benchmarking values, it is important to look at comparable outcomes from a variety of sites. The NurseMidwifery Clinical Data Set (NMCDS) was developed to “describe, compare, and assess the quality of nursemidwifery care in all possible birth settings, including hospitals, birth centers, and homes” (Greener, 1991, p. 174) and describes the structure, process, and outcomes of nurse-midwifery care. Although earlier studies concluded that birth centers were a safe practice site (Albers & Katz, 1991; Fullerton et al., 1997; Rooks et al., 1989), it remains vital for birth centers and certified nurse-midwives (CNM) to continue to report clinical outcomes. The NMCDS can provide a basis for nursemidwifery outcomes research (Greener, 1991; Jackson, Lang, Dickinson, & Fullerton, 1994). The majority of the data points used in the automated birth log reported in this article were matched to data points in the NMCDS.

Study Setting BCHC is a multidisciplinary nursing health care clinic and freestanding birth center that opened in April 1994 as a joint venture between the University of Utah College of Nursing and the University of Utah Hospital and Clinics to provide obstetric, gynecologic, pediatric, and mental health services. CNM faculty members provide gynecologic, antepartum, intrapartum, and postpartum care. Women who are expected to have an uncomplicated normal spontaneous vaginal delivery are offered labor, delivery, and immediate postpartum care for mother and newborn in the freestanding birth center. Exclusion and transfer criteria are shared with the woman and her partner at the first obstetric visit. Antepartum exclusion criteria include multiple gestation, malpresentation, previous cesarean delivery, oligohydramnios, polyhydramnios, intrauterine growth retardation, drug abuse, less than 37 weeks gestation, more than 42 weeks gestation, pregnancy-induced hypertension, severe anemia, active herpes lesion, placenta previa, known birth defect, and any maternal medical complication or chronic disease. When women enter the birth center for labor assessment, those who are not in active labor are sent home to return when active labor is achieved. Women who are in active labor and continue to expect an uncom-

January I February 2002

plicated vaginal delivery are admitted to the birth center. If the labor assessment reveals potential problems with the mother or the baby, or if complications develop after admission to the birth center, the mother and/ or baby are transferred to the University of Utah Hospital via private car or ambulance. Intrapartum exclusion and transfer criteria include rupture of membranes longer than 12 hours without active labor; pregnancyinduced hypertension; prolapsed cord; thick meconium-stained fluid; evidence of chorioamnionitis or other active infectious process; fetal malpresentation; arrest of labor; fetal distress; abnormal bleeding (placenta previa or abruptio); patient desire; and any medical, surgical, or obstetric development requiring more complex treatment than is available at the birth center. Postpartum transfer criteria include uncontrolled postpartum hemorrhage or severe controlled hemorrhage requiring a blood transfusion, pregnancy-induced hypertension, retained placenta (more than 1 hour), postpartum infection, and lacerations requiring repair in a hospital. Newborn transfer criteria include cardiac or respiratory distress, major congenital anomaly, seizure activity, and persistent hypothermia or hyperthermia. Women and their partners were expected to attend a 36-week birth center class in which all exclusion and transfer criteria were reviewed and given again in writing. Logistics of how and when to call the CNM, what to bring to the birth center, and other questions were answered in the 36-week class. At the time of this study, each time a woman received a labor assessment in the birth center, the visit was recorded on her paper chart. The registered nurse or CNM then summarized highlights of the encounter in the birth log. The paper birth log consisted of ll-inch by 17-inch pages collected in a large loose-leaf binder. The BCHC birth log served as the federally mandated delivery record, a source of information for clinical research and quality improvement studies, and a source of information for management and business activities. Birth log data were extracted at intervals to create reports. When information was required from the birth log, the clinic manager or a CNM reviewed the paper birth log and manually tabulated the information. For larger reports, data from the birth log were sent to the college of nursing to be transcribed, and standard reports were generated and returned to the clinic. However, the reporting process was time-consuming, and hand calculations were cumbersome and labor intensive. It could take several hours to tabulate data for a single report, and routine reports from the college of nursing would take weeks to return to the clinic. Difficulty obtaining timely reports was consistently identified by the CNMs at BCHC as a high-priority problem.

JOGNN 111

The CNMs had also identified flexible formatting as a highly desirable feature in a reporting system because in many cases the same data must be reported to different entities in a slightly different format. Because the birth log is a source of data for research and quality improvement studies, CNM faculty at BCHC needed to be able to obtain a variety of reports in addition to the standard reports generated by the college of nursing.

Automated Technology Automated technology can assist the provider in efficiently collecting and managing information for outcome studies, quality improvement, marketing, planning, and other purposes that support professional nurse-midwife practice (Bowles, 1997; Johnson, 1994; McDaniel, 1997). Unfortunately, many of the existing health care databases do not adequately capture nursing outcome data (Bowles, 1997; McDaniel, 1997; Zielstorff, 1995). Although there is already an automated system based on NMCDS, called Gravidata (Jackson et al., 1994), clinicians at BCHC needed additional, institution-specific data not included in the Gravidata database, such as data on nursing students and data for business processes. In evaluating data needed for an electronic medical record, Henry and Mead (1997) found that the nursing classification systems they evaluated were insufficient for recording all pertinent nursing data in an electronic medical record. In a similar manner, the NMCDS was found to be comprehensive but insufficient for capturing all of the data required by the BCHC clinic. Prototype systems, like the one described in this study, help clinicians define what they need and want in an electronic record. It is vital that nurses participate both in defining benchmarking data for nursing practice and in developing systems that collect nursing clinical outcomes data. In 1993, the American Nurses Association identified informatics as a nursing specialty that focuses on information handling. Nurses who practice informatics at the generalist level are registered nurses, usually with a bachelor's degree, who have additional knowledge and experience; an informatics nurse specialist requires a master's degree and is in advanced nursing practice (American Nurses Association, 1995). The additional knowledge and experience needed by informatics nurses include computer literacy, programming skills, and a knowledge of the following: systems analysis and design, database development, the interaction between people and technology (human factors), systems implementation and support, and testing and evaluation of health care applications. Informatics nursing involves communications skills and liaison activities between engineers, analysts, developers, and health care system

112 JOG"

users. Informatics nurses practice collaboratively with computer systems professionals and health care professionals who have expertise in a variety of clinical settings. The CNMs at BCHC collaborated with informatics nurse specialists and informatics students from the college of nursing to design a prototype system to document clinical outcome data and measure key quality indicators. The data elements are founded in part on the NMCDS. Although the database is a prototype intended for eventual integration into the electronic medical record, it also assists the CNMs with outcomes reporting and communication while the electronic medical record is being developed.

U i f f i c u l t y obtaining timely reports was consistently identified by the CNMs at Birthcare Healthcare as a

high-priority problem.

As a prototype, the database was frequently revised. The initial phase automated the birth log; later phases expanded the database. This article reports a sample of the clinical outcomes obtained at the BCHC birth center and describes the automated birth log prototype database used to manage the outcome reports.

Method Database Development The initial database was designed and tested for correct and consistent design and data handling during informatics students' graduate coursework. The prototype database was constructed in Microsoft Access because of its availability to the clinicians and its ease of use. A graphical interface was developed and evaluated with sample data; the clinicians at BCHC found the system easy to learn and use (Sward, 1998). The database records demographic data, visits to the birth center, and selected clinical outcomes. The user interface was designed to facilitate data entry. Whenever possible, data are entered by selecting items from a list instead of typing. This helps to prevent spelling inconsistencies and standardizes data collection. Most of the lists are developed using data points from the NMCDS. Internal rules check certain data as they are entered. For example, only whole numbers between 0 and 10 can be entered as Apgar scores.

Volume 30, Number 1

FIGURE 1

Data entry form. Copyright by the University of Utah College of Nursing. Reprinted with permission.

Data entry forms are organized conceptually, with sections such as demographic information, social factors, laboratory results, and visits to the birth center. Data are also organized chronologically, with maternal demographic information first, followed by information specific to this pregnancy, and then labor, delivery, and newborn information. A tab design provides the familiar look of the paper chart. Figure 1 shows a sample data entry screen. Standardized queries and reports were developed and evaluated for accuracy of the data reported (Sward, 1998). Clinicians can select the standardized queries and reports from a simple menu system. Queries and reports can be retrieved for any interval, and in addition, the CNMs can create any number of ad hoc queries. Figure 2 shows the query menu. Figure 3 shows two of the standard reports.

Study Methods The design of the database was evaluated during graduate coursework, and the user interface, standard queries, and reports were evaluated using data from imaginary clients. Database queries were compared to hand-tabulated reports of the same data to verify the query design. A retrospective chart review was used to verify that the database design was appropriate for actual client January / February 2001

data. Because the database is a prototype and had not been widely implemented at the time of the study, we entered all data for this study. After institutional review board approval was obtained, the birth log was reviewed for a 2-year period from July 1, 1997, to June 30, 1999, and data were entered into the database. We generated clinical outcome reports for each quarter of each calendar year, for each year, and for the 2-year study period. Calculations were performed within the database as part of the queries.

Selection of Outcomes Clinical outcome data selected for this study include the number of deliveries a t the birth center, the rate of transfer to the hospital, cause-specific transfer rates, rates of meconium-stained fluid, type of medications in labor, number and percentage of low Apgar scores, and low and high birth weights. The length of stay and method of transport provide institution-specific data for evaluating cost-effectiveness. Because BCHC is licensed as an outpatient facility for stays of less than 24 hours, length of stay was calculated for the mothers from admission (the time the labor assessment began) to discharge, admission to birth, and birth to discharge (see Figure 4). The birth center policies allow for a 12hour postpartum stay. However, women who deliver in the afternoon are allowed to remain at the birth

JOGNN 113

FIGURE 2

Query menu. Copyright by the University of Utah College of Nursing. Reprinted with permission.

Admissions, Births, a d Transfers 220

2CQ

19

8

91-82%

711197 to 6Bl#9

8.64%

3.64%

FIGURE 3

Reports. Copyright by the University of Utah College of Nursing. Reprinted with permission.

224 JOG"

Volume 30, Number 1

Times By Month 25.00

20.00

-

15.00

? o

~

f.

Birth to Discharge 8 Admission to Birth

~

E

F 10.00

5.00

0.00

MonthlYear

FIGURE 4

Average times of length of stay by month (admission to birth and birth to discharge).

center through the night and are discharged the next morning. The clinical outcomes reported are key quality indicators commonly found in the literature. Additional outcome measures that are part of the database but have not been reported here include the number of deliveries in which students participated, neonatal intensive-care unit admission rate, newborn and postpartum readmission rate, perineal lacerations, laboratory tests such as RH status and the number of clients who have positive group-B strep tests, use of the Jacuzzi tub in labor, and postpartum functional status of mother and newborn.

Results Data entry took approximately 5 minutes per birth. As data were entered, occasional minor changes were required in the database structures. These changes were primarily to increase the size of a field when the actual text, such as a woman's last name, was longer than the database allowed. Content changes involved additions to the medication list; individual medications were then matched to the corresponding category from the NMCDS. Queries to determine demographic characteristics of the study population were added to the standard queries. January / February 2001

The previously designed queries were modified, primarily with an adjustment to the groupings. For example, the age query originally was designed to show women younger than age 18 in the first group; women who were 18 were placed in the second group. To facilitate comparison with the NBCS (Rooks et al., 1989), the groups were modified so that 18-year-olds were in the first group. In a similar manner, the count by parity was modified to isolate the women with a parity of one. The only other changes to queries were formatting issues, such as grouping the results of multiple queries into a single table.

Demographic Characteristics In the 2-year period, 231 women were evaluated in the birth center, for a total of 269 labor assessments (some women were assessed more than once). Table 1 shows the client characteristics. Of the women assessed, 43 (16%) were sent home because they were not in active labor, 6 YO) were admitted directly to University Hospital, and 220 (82Y0)were admitted to the birth center. Of the 220 women admitted to the birth center, 202 (92%) delivered at the birth center. The BCHC study population is approximately 96% white non-Hispanic. The majority of the study population, 182 (83%), were between the ages of 19 and 35.

JOG"

11.5

Twenty-four (11%)were older than 35, and 5 (2%)were 18 years of age or younger. Of the women admitted to the birth center, 72 (33%) were nulliparous, 72 (33%) had a parity of one, 65 (30%) had a parity of two to four, and 9 (4%) had a parity of five or more.

Medication Use Of the 220 women admitted to the birth center, 39 (18%) women received 46 medications in labor. Narcotics were given to 36 (16%) of the women, making up 78% of the medications; antibiotics were given to 4 (2%) of the women, making up 9% of the medications; antiemetics were given to 3 (1%) of the women, making up 7% of the medications. Other medications given were 1 each (0.5%) of oxytocin, methylergonovinemaleate, and acetaminophen, each making 2% of all medication given in labor.

Intraparturn, Postpartum, and Newborn Transfers Of the 220 women admitted to the center, 19 (8.6%) were transferred to the hospital. Eighteen women (8.2%) were transferred intrapartum and 1 (0.5%) was transferred postpartum for a retained placenta (see Table 2). For the intrapartum transfers, 10 (4.5% of the 220 women) were for arrest of labor, 4 (1.8%) desired epidural anesthesia, 6 (3%) required assisted delivery, and 1 (0.5%) had meconium-stained fluid. There were 8

TABLE 1

Client Characteristics Characteristic

Number

Percentuge

Labor assessment outcome A: Home, not in active labor B: Admit to hospital C:'Admit to birth center Tnt.! n?unhpr nf assessments Number of births Number maternal transfers Number newborn transfers Age Younger than or equal to 18

19 to 35 Older than 35 Unspecified Parity 0

1 2 to 4 Greater than 4

116 JOG"

43 6

220

16 2 82

269

202 19 8

91.82 8.64 3.64

(4% of 220) newborn transfers. One (0.50/,) was for meconium aspiration, 3 (1.5%) for transient tachypnea of the newborn, and 5 (2.5%) for respiratory distress. Four of the transferred newborns had a 5-minute Apgar score of less than or equal to 7. Of the 220 women admitted to the birth center and the 18 who were transferred to the hospital, 2 (1%) were delivered by primary cesarean delivery, 2 (1%) by vacuum extraction, and 2 (1%) by low forceps. Interestingly, 66% of the newborns of transferred women were male, whereas in the overall birth center population, 52% of newborns were male. Methods of transport can include private car, ambulance, and air transport. The ambulance was called for 6 newborn transfers (2.9% of all newborns, or 75% of newborn transfers) and for 4 maternal transfers (1.8% of the 220 women admitted, or 21% of the maternal transfers). Private cars were used for all other transfers. No air transport was required.

Newborn Data Newborn outcomes for the 202 birth center births are shown in Table 3. The gestational ages of infants were 1 (0.5%) at 37 weeks, 195 (96.5%) at 38-41 weeks, and 6 (3%) at 42+ weeks. Two newborns had a birth weight of less than 2,500 g (1%). The incidence of babies weighing more than 4,000 g was 28 (13.9%). The incidence of a 1-minute Apgar score of less than or equal to 5 was 1 3 (6%), and a 5-minute Apgar score less than or equal to 7 was 7 (3%). Of these 7 infants, 4 ( 5 7 % )were transferred because of respiratory distress.

Length of Stay Over the 2-year period of this study, time from admission to discharge averaged 15 hours and 50 minutes (with a standard deviation of 5 hours, 51 minutes), time from admission to birth averaged 3 hours and 59 minutes (with a standard deviation of 3 hours, 32 minutes), and time from birth to discharge averaged 12 hours and 35 minutes (with a standard deviation of 3 hours, 31 minutes).

Discussion 5

182 24 9 72 72 65

9

2 83 11 4 33 33 30 4

Clinical Outcomes As noted previously, although data from other studies may provide a frame of reference, each setting has unique factors and should have unique benchmarking criteria (Fullerton et al., 1997). BCHC clients tend to be a fairly homogeneous, low-risk group who self-select to nurse-midwifery care. As was found in previous studies (Albers & Katz, 1991; Fullerton et al., 1997; Jackson et al., 1994; Rooks et al., 1989), most of the women are

Volume 30, Number 1

TABLE 2

Transfer Information Characteristic

Number

Percentage

18 1 8

8.2 0.5 3.64

4 15 6 2

1.8 6.8 2.9 1

2 2 2 12

0.9 0.9 0.9 5.4

10 4 1 2 1 2

4.5 1.8 0.5 0.9 0.5 0.9

5 3 1

2.5 1.5

Transfer type Maternal intrapartum Maternal postpartum Newborn Transfer method Ambulance (maternal) Car (maternal) Ambulance (newborn) Car (newborn) Intrapartum transfer outcomes Primary cesarean delivery Low forceps Vacuum extraction Normal spontaneous vaginal delivery Maternal transfer by cause. Arrest of labor/failure to progress Epidural desired Meconium-stained amniotic fluid Forceps/assisted delivery needed Retained placenta Other/unspecified Newborn transfer by cause' Respiratory distress Transient tachypnea Meconium aspiration

0.5

Comment

Also listed with arrest of labor

Also listed with respiratory distress

a Some women and newborns had more than one problem listed as the reason for transfer.

TABLE 3

Birth Weight, Gestational Age, and Apgar Scores Characteristic

Gestational age of infants (weeks) Less than or equal to 37 38 to 41 42+ Birth weight of infants (g) Less than 2,501 2,501 to 4,000 Greater than 4,000 Apgar scores 1-minute Apgar less than or equal to 5 5-minute Apgar less than or equal to 7

Number

Percentage

1 195 6

0.5 96.5 3.0

2 172 28

1.o 85.1 13.9

13 7

0.06 0.03

Comment

1 transfer to the hospital for transient tachypnea

4 of these 7 were transferred to the hospital for

respiratory distress

January f February 2001

JOG"

117

in the 18- to 35-year-old age group, and most are nulliparous or primiparous. The lack of ethnic diversity, although somewhat reflective of the population in Utah, makes it difficult to generalize the outcomes to another population. Clinical outcomes reported in this study are consistent with previous studies (Fullerton et al., 1997; Rooks et al., 1989) that support the use of a freestanding birth center as a safe alternative delivery site. The maternal transfer rate at BCHC (8.6%) was lower than the 15% to 24% maternal transfer rate found in the literature (Fullerton et al., 1997; Jackson et al., 1994; Rooks et al., 1989). The CNMs at BCHC adhere stringently to the admission criteria previously described. Of the 23 1 women evaluated at BCHC, 6 (2%) were directly admitted to the hospital. Although these women desired a birth center birth, they are not counted in the overall transfer rate. Had the women been admitted to the birth center and subsequently transferred to the hospital, the transfer rate would have been higher. At BCHC, 18% of admitted women received medications in labor. In the NBCS, 24% of nulliparous women and 6.2% of parous women received labor medications. The NBCS data showed a 4.4% cesarean delivery rate, whereas BCHC had a 0.9% primary cesarean delivery rate for women admitted to the birth center. If the number of women who were directly admitted to the hospital (6) were included, the cesarean delivery rate would be slightly higher. Newborn outcomes at BCHC were similar to the NBCS data and reflected the CNMs' adherence to admission criteria. For example, the NBCS showed that 86.2% of newborns were born at between 38 and 41 weeks gestation; at BCHC, 96.5% of newborns were born in this gestational range. In the NBCS, 0.8% of newborns had a birth weight below 2,501 g compared with 1.0% at BCHC. Length of stay data can help assess the need for RN coverage and assist in budgeting efforts. Women who delivered at BCHC knew that their length of stay could be up to 12 hours. Should a longer stay be medically indicated, the woman could stay up to 23 hours and 59 minutes as an outpatient at the birth center, and insurance was billed for additional observation time. Alternatively, if a woman desired a longer recovery time, but it was not medically indicated, she could pay an hourly rate for additional hours of recovery service. In the 2 years of this study, the time from birth to discharge averaged 12 hours and 35 minutes. One of the most impressive clinical outcomes is the time from admission to birth, with a 2-year average of 3 hours and 59 minutes. The nurse-midwifery philosophy of low-risk obstetric care is exemplified here. Women were not admitted until they were in active labor,

118 JOG"

and they were encouraged to ambulate freely, use the Jacuzzi tub and birthing ball, and take oral fluids and light foods. On admission, women were placed on the fetal monitor until a reactive tracing was obtained; thereafter, intermittent Doppler fetal assessment was used. When the woman was ready to push, she could assume a variety of positions until she found the most comfortable and effective position.

k l i n i c a l outcomes of this study, such as a

1% cesarean delivery rate, illustrate the nurse-midwifery philosophy of low-risk obstetric care.

Study Design and Methods Methodological concerns with this study are similar to those of previous birth center studies (Albers & Katz, 1991; Fullerton et al., 1997; Jackson et al., 1994; Rooks et al., 1989). This descriptive study had no comparison group, a relatively small size, and used retrospective data. Jackson et al. (1994) noted that the lack of a control group prevents comparison between studies and that birth centers have varying admission and exclusion criteria. Although this study used data from all women seen in a 2-year period, the overall size of the study was only 220 women, which may be too small to investigate outcomes relating to rare events. The use of retrospective data has long been seen as problematic but difficult to avoid in birth center studies (Albers & Katz, 1991). Retrospective data issues include determining whether an item is defined and measured in the source (in this case, the paper chart by way of the birth log) the same way it is defined and used in the destination (the database) and missing data. Weights of 3 newborns were missing, as were the ages of 9 mothers. Of the 269 labor assessments, 15 records had missing or questionable admission or discharge dates or times. In these cases, we used the chart progress notes to obtain required information. The records with missing data were omitted from the queries that could potentially affect the results. In this study, data were transcribed from the birth log rather than from the paper charts, adding to the potential for transcription error. CNMs and experienced registered nurses summarized data from the charts in the birth log, and we abstracted data from the birth log into the database. Jackson et al. (1994) suggested that using the actual care providers for data classifiers is

Volume 30, Number 1

optimal because they are familiar with both obstetric care in general and with the criteria and practices of the specific setting. In this test of the prototype’s ability to document and retrieve data in a real practice setting, it was helpful to have an experienced CNM evaluate the clinical data and an informatics nurse deal with any changes that needed to be made in the database. The clinicians will eventually enter data directly into the electronic medical record. The need to transcribe from one source to another will be eliminated. We found that all the desired data could be entered into the database. Once the design of the queries and grouping ranges was determined, changing the reports for a variety of date ranges was a simple matter of entering the beginning and ending dates on the menu form and running the queries. A process that previously had taken several hours had been reduced to a few minutes.

Summary and Conclusions Clinical outcomes in this study are consistent with previous research of birth center outcomes that demonstrate that birth centers are a safe alternative site for delivery. Future birth center outcomes research should examine the relationship between nulliparity and the transfer rate for epidural anesthesia, between a 5-minute Apgar score of less than 7 and the newborn transport rate, and between infant gender and antepartum maternal transfer rate. CNMs need to report a variety of structure, process, and clinical outcomes. Continuous quality improvement and clinical practice research require the ability to evaluate aggregate client data. Information on room use and length of stay assist with budgeting and other management decisions. Information such as cesarean delivery rate, readmission rate, and the maternal transfer rate can assist with contract negotiations with insurance companies. Outcomes reporting can be optimized with automated technology. A system such as the automated birth log described here provides a simple and efficient way to collect and report desired outcome data. When regulatory bodies o r state agencies request practice statistics such as adherence to mandatory PKU testing, this information is available immediately on querying. Reports can be easily generated for review by peers, hospitals, insurers, and group health plans. The automated birth log described here is not only a mechanism for generating outcome reports, it is also a communications tool. For example, the newborn procedure form lists routine procedures such as infant vitamin K injections and hearing screenings. This form not only documents those activities but also serves as a

January /February 2001

reminder to the nurses to perform the activity. The automated birth log can also help practitioners manage their workload. Among the standard queries are lists of clients arranged in various ways, such as by due date and by expected delivery site. The automated birth log can generate a list of clients who have not returned for their final postpartum examination.

O u t c o m e s reporting can be optimized with automated technology.

The graphical interface is user friendly and contains data validation features. Data are selected from a list whenever possible to minimize data entry errors. An important feature is the compliance with the NMCDS, which promotes compatibility of data from this database with data from other systems. Automated technology provides a simple and efficient way to collect and report data and should be used by advanced practice nurses to determine regional and national benchmarking data. Information on a national level regarding CNM processes and outcomes will help strengthen the position of nurse-midwifery care in the health care industry today and in the future.

REFERENCES Albers, L. L., & Katz, V. L. (1991).Birth setting for low-risk pregnancies: Analysis of the current literature. Journal of Nurse-Midwifery, 36(4), 215-219. American Nurses Association. (1995). Scope of practice for nursing informatics. Washington, DC: American Nurses Publishing. Bowles, K. H. (1997). The barriers and benefits of nursing information systems. Computers in Nursing, 15(4), 191-196. Fullerton, J. T., Jackson, D., Snell, B. J., Bessner, M., Dickinson, C., & Garite, T. (1997). Transfer rates from freestanding birth centers: A comparison with the National Birth Center study. Journal of Nurse-Midwifery, 42(1), 9-16. Greener, D. (1991). Development and validation of the Nurse-Midwifery Clinical Data Set. Journal o f NurseMidwifery, 36, 174-183. Henry, S., & Mead, C. (1997).Nursing classification systems: Necessary but not sufficient for representing “what nurses do” for inclusion in computer-based patient record systems. Journal of the American Medical Informatics Association, 4, 222-232. Jackson, D., Lang, J. M., Dickinson, C. P., & Fullerton, J. T. (1994). Use of the Nurse-Midwifery Clinical Data Set

lOGNN 119

for classification of subjects in birth center research. Journal of Nurse-Midwifery, 39(4), 197-213. Johnson, D. (1994). Automation of the birth registry log book. Computers in Nursing, 12, 245-252. McDaniel, A. M. (1997). Developing and testing a prototype patients care database. Computers in Nursing, 25(3), 129-136. Rooks, J., Weatherby, N., Ernst, E., Stapelton, S., Rosen, D., & Resenfield, A. (1989). The National Birth Center Study. New EnglandJournal of Medicine, 321, 1804-1811. Sward, K. A. (1998). Database usability: Automation of the Birthcare Healthcare birth log. Unpublished master's project. University of Utah, Salt Lake City. Zielstorff, R. D. (1995). Capturing and using clinical outcomes data: Implications for information systems design. Journal of the American Medical Informatics Association, 2(3), 191-195.

Leissa Roberts is the clinical director at Birthcare HealthCare and at OB Care Now in Salt Lake City, UT.In addition to serving as the clinical directov, she is a practicing certified nurse-midwife in both programs and an assistant professor (clinical) at the University of Utah College of Nursing, Salt Lake City. Katherine Sward received a master's degree in nursing/clinical informatics from the University of Utah. She a computer professional and clinical instructor at the University of Utah College of Nursing, Salt Lake City. Address for correspondence: Leissa Roberts, RNC, MS, CNM, University of Utah, College of Nursing, 10 S. 2000 E. Front, Salt Lake City, UT 84112-5880.

Index to Advertisers Agilent Technologies Asante Health System AWHONN Bry anlgh Dale Medical Products Franciscan Skemp Healthcare Integrity Pharmaceuticals Ortho-McNeil Pharmaceutical, Inc. Sentara Healthcare SkyTron UNC Hospitals University of Cincinnati University of New Hampshire University of Texas at Austin Virginia Commonwealth University Yukon-Kuskokwim Health Corp., Inc.

2 2 124,125 Cover 3

4 8 7-8 Cover 2-1 Cover 3 Cover 4 2 10 10 10 8 10

This index is provided as a service. The publisher does not assume any liability for errors or omissions.

120 lOGNN

Volume 30, Number 1