Lessons For Providers And Hospitals From Philadelphia’s Obstetric Services Closures And Consolidations, 1997–2012

Abstract

The Affordable Care Act is triggering an increase in hospital consolidation and mergers. How other hospitals respond to these disruptions in supply could influence patient outcomes. We examined the experience of Philadelphia County, Pennsylvania (coterminous with the city of Philadelphia), where thirteen of nineteen hospital obstetric units closed between 1997 and 2012. Between October 2011 and January 2012 we interviewed twenty-three key informants from eleven hospitals (six urban and five suburban) whose obstetric units remained open, to understand how the large number of closures affected their operations. Informants reported having confronted numerous challenges as a result of the obstetric unit closures, including sharp surges in delivery volume and an increase in the proportion of patients with public insurance or no insurance. Informants reported adopting a number of strategies, such as innovative staffing models, to cope with the added demand brought about by the closure of nearby obstetric units. Informants emphasized that interhospital communication could mitigate closures’ stresses on the health care system. Our study supports the need for policy makers to anticipate reductions in supply and monitor patient outcomes.

During the past thirty years the number of hospital consolidations, closures, and mergers has fluctuated, with a spike in the mid-1990s that coincided with an increase in managed care insurance plans.¹ Hospital consolidations, closures, and mergers have accelerated since the passage of the Affordable Care Act (ACA): An average of 89–98 mergers per year were reported between 2011 and 2013, compared to an average of 30–60 mergers per year between 2002 and 2010.^1–5

Studies have shown that these consolidations may affect prices charged to consumers^6–8 for hospital services and increase racial/ethnic and income disparities in access to various health care services.⁹ The impact of these consolidations on patient health is less clear. One study showed negative effects for patients with acute time-sensitive conditions such as acute myocardial infarction and trauma.¹⁰ Other studies showed improved outcomes, perhaps because higher delivery volumes improved health care providers’ comfort in dealing with sick infants^11,12 or improved the coordination of care.¹

There are more than four million births annually in the United States, and childbirth is the leading reason why women ages 18–44 are hospitalized in the country.¹³ Thus, ensuring adequate access to obstetric health care is an important public health issue.

We examined the case of Philadelphia County, Pennsylvania (which is coterminous with the city of Philadelphia). In that county, between 1997 and 2012, thirteen of nineteen obstetric units stopped providing obstetric services. These closures resulted in a reduction in the number of available obstetric beds in the county by approximately 50 percent. Hospital obstetric units that remained open in the region experienced an average increase in delivery volume of 57.7 percent during the same period (calculated from data in Exhibit 1), according to birth certificate records and the Maternity Care Coalition.¹⁴ These changes occurred with no real change in the total number of deliveries in the five-county Philadelphia region in the same time period.

Exhibit 1.

Changes In Hospitals’ Delivery Volume And Insurance Status In The Philadelphia Region, By County, 1995–2009

County	Hospital	Status	1995 Deliveries	2009 Deliveries	% Change Deliveries	% Change in Fed/Uninsured
Bucks	A	Open	1,051	1,363	29.69	247.72
	B	Open	1,107	1,596	44.17	61.39
	C	Open	1,390	1,128	−18.85	152.38
	D	Open	1,414	1,920	35.79	114.65
	E	Hospital Closed	142
Chester	F	Open	2,042	2,567	25.71	94.6
	G	Open		418
	H	Open	296	2,280	670.27	−30.76
	I	Open	1,209	924	−23.57	299.8
	J	Unit Closed	760
Delaware	K	Open	2,603	2,107	−19.05	447.26
	L	Open	1,226	1,657	35.15	987.16
	M	Open	814	833	2.33	199.02
	N	Unit Closed	1,319
Montgomery	O	Open	3,243	5,155	58.96	76.38
	P	Open	2,021	1,915	−5.24	−15.16
	Q	Open	1,791	2,770	54.66	746.7
	R	Open	1,714	2,666	55.54	57.91
	S	Open	674	776	15.13	309.14
	T	Open	773	694	−10.22	21.21
	U	Unit Closed	812
	V	Unit Closed	565
Philadelphia	W	Open	1,507	2,932	94.56	142.1
	X	Open	1,597	2,081	30.31	321.58
	Y	Open	2,307	4,004	73.56	110.99
	Z	Open	3,585	4,790	33.61	268.98
	AA	Open	1,644	3,171	92.88	122.45
	AB	Open	2,340	2,174	−7.09	21.3
	AC	Hospital Closed	476
	AD	Hospital Closed	1,034
	AE	Hospital Closed	967
	AF	Hospital Closed	844
	AG	Hospital Closed	771
	AH	Unit Closed	655
	AI	Unit Closed	279
	AJ	Unit Closed	1,264
	AK	Unit Closed	330
	AL	Unit Closed	1,208
	AM	Unit Closed	1,865
	AN	Unit Closed	711
	AO	Unit Closed	1,144

SOURCE Authors’ analysis of birth certificate records, Pennsylvania Department of Vital Statistics. NOTE Data confidentiality agreements precluded listing hospitals by name.

This degree of obstetric unit closures was larger than that experienced by any other metropolitan area. However, the trend is consistent with the results of one previous study showing that California hospitals often eliminate obstetric units when they reduce services.¹⁵ The Philadelphia County hospitals that closed obstetric units tended to have smaller delivery volumes, to have no academic affiliation or obstetrics training program, and to be located in the northeastern or northwestern areas of the county.¹⁶

The closures of Philadelphia obstetric units were initially associated with an increase in neonatal mortality, but that abated over time.¹⁶ The sustained reductions in obstetric units in Philadelphia and the observed adverse outcomes provide an opportunity to understand more broadly how hospitals respond to closures of nearby clinical units or entire hospitals, regardless of the reason for the closure. This is an important issue for any community or public health department.

We conducted a qualitative study of key informants involved in obstetrical care in Philadelphia to answer several questions. First, what is the perceived effect of an acute reduction in obstetric care services on the hospitals that remain open? Second, what changes do hospitals and providers in remaining hospitals make in response to changes in patterns of care? Third, what opportunities do obstetric providers identify to mitigate the adverse effects of changes in access to obstetric care?

We hypothesized that our informants would consistently identify changes in delivery volume and patient characteristics, both medical and sociodemographic, as important challenges with closures of obstetric units. Additionally, we sought to understand the role that a local public health department or a regional non-governmental organization could play in responding to these changes.

Study Data And Methods

SUBJECTS

Between October 2011 and January 2012 we conducted semistructured interviews with twenty-three key informants at six urban hospitals and five suburban hospitals in the greater Philadelphia area. The obstetric units in these hospitals were still open as of the beginning of 2012 and had remained open as of October 2014.

The hospitals were chosen based on their location relative to Philadelphia County and their change in delivery volume since 1995–96. Thirteen informants were employed at one of the six remaining Philadelphia County obstetric units, and the other ten were employed at one of five selected units in the four counties that border Philadelphia.

To build our pool of informants, we first recruited obstetric department chairpersons and leaders of private obstetric groups that delivered babies at the eleven study hospitals, because these people were directly affected by the closures of obstetric units in Philadelphia County. Then, based on their suggestions, we identified and recruited additional clinicians at each study hospital, including obstetricians, nurses, nurse managers, and midwives. Each informant had been at his or her respective hospital for at least ten years and had been practicing obstetrics for 10–48 years.

Saturation—the point at which no new or relevant information emerged from further interviews—was reached with a total of twenty-three informants. There were eight chairpersons, eight obstetricians, three nurses, three nurse managers, and one midwife, with one or two physicians and one or no nurse from each of the eleven hospitals.

At the time of the interviews, tacit verbal consent was obtained from each informant. The Institutional Review Board at the Children’s Hospital of Philadelphia approved all study procedures. The Institutional Review Board did not allow us to identify informants by title or hospital, because doing so would breach participants’ confidentiality and result in the identification of the source of several quotations in the article.

INTERVIEW METHODS

A public health professional trained in interviewing techniques conducted the in-person and telephone interviews. Each informant chose the interview method he or she preferred.

Each interview was digitally recorded and lasted 30–90 minutes. The interviewer used a semistructured open-ended interview guide that included questions about the response of the informant’s hospital and obstetric unit to the closures of obstetric units in the area, the biggest challenges faced by hospitals and obstetricians, the impact of the closures on the financial health of the informant’s obstetric unit, and suggestions for how hospital units that remain open can plan for and respond effectively to future closures.

ANALYSIS

The interviews were transcribed and edited to remove identifying information. The transcripts were imported into Atlas.ti, version 6.2—a qualitative data analysis software package used for coding—and a codebook was developed using grounded theory.^17,18 Two experienced independent coders separately analyzed the data and then compared findings to reduce the chance of errors from coding and increase the reliability of the themes identified in the data.

The themes presented in the Study Results section were not chosen a priori but were developed from informants’ responses. We assessed the reliability and validity of the analyses through four criteria described in previous research: prolonged engagement with the data, credibility, transferability, and confirmability.¹⁹ With the exception of some minor edits to improve clarity, all quotes are provided verbatim.

LIMITATIONS

Our study had several limitations. First, this project was not designed to assess changes in patient satisfaction resulting from closures.

Second, there are elements of the perinatal system in Philadelphia that may occur in only a few areas of the country, such as women receiving care at hospital clinics. These elements may limit the generalizability of our results about the prenatal care system to other areas where clinicians in private practice deliver the majority of prenatal care.

Third, the ability of the local public health department to intervene within the obstetric care system that we found in Philadelphia might not exist in other regions. Thus, they may need other solutions, such as regional hospital collaborations outside of the governmental system.

Finally, because this was a qualitative study, we could not use our results to show a causal association between any changes in response to the closures in Philadelphia that were made by the obstetric units that remained open and any change in perinatal outcomes. Nor could we assess whether the decision to participate in this study was related to either positive or negative experiences surrounding the closures. However, no informant gave information that was entirely negative or positive during his or her interview.

Study Results

Informants generally described a severely strained obstetric care system following the widespread closures of units or hospitals in the region. In most cases, informants described the closures as happening abruptly. Other hospitals “did not have a lot of warning,” one informant said, and thus each closure precipitated immediate challenges in the remaining obstetric units.

Several themes and subthemes emerged from our interviews (Exhibit 2) that confirmed data presented in Exhibit 1. First, informants identified economic challenges in obstetrics as a major underlying reason for obstetric unit closures and ongoing challenges within open units.

Exhibit 2.

Themes And Subthemes In Philadelphia Region Obstetric Unit Closures, Identified By Twenty-Three Key Informants

Reasons for Closures	System Stressors
Liability	Volumes Surges
Poor Reimbursement	Discontinuity of Prenatal Care
Minimal Finances	Patient Mix Changes
	Staffing Challenges
Responses to Closures	Prevention of Adverse Effects
Staffing Model Changes	Region-Wide Protocols
Structural Changes	Communication Between Hospitals

SOURCE Authors’ analysis.

Second, informants described both acute and chronic stresses to the system that arose from changes in patient volume and mix, precipitated by the closures, in the remaining obstetric units. Informants said that they were able to adapt to some of these stresses but not others.

Third, informants described an ad hoc response to obstetric unit closures with little advance planning. The response was reactive, based on changing volumes and patient mix.

Fourth, informants offered suggestions for preventing adverse effects from future closures.

REASONS FOR CLOSURES OF OBSTETRIC UNITS

Our informants identified the low financial margins and high fixed costs of running obstetric units as primary reasons for closures. “In Philly, you get paid less, it costs more, and the liability is higher. So ‘perfect storm’ is an overused phrase, but it’s at least a triple whammy,” one person said. Informants identified several unique factors related to obstetrics that led to high fixed costs: high community malpractice rates, which in turn reinforced the need for hospitals to employ obstetricians directly; and high overall personnel costs to run a labor and delivery floor.

Informants reported that payments did not keep pace with these expenses. This fact was exacerbated by the high percentage of women with public or no insurance who had been cared for by obstetric units immediately before they closed (Exhibit 3), who might have had lower reimbursement for services compared to women with private insurance.

Exhibit 3.

Percentage Of Deliveries With Public Or No Insurance At Closed Obstetric Units In The Philadelphia Region, 1995 And Last Year Open

Hospital	% Deliveries with Federal or no insurance, 1995	% Deliveries with Federal or no insurance, last open year
E	14.80%	12.30%
J	16.80%	52.20%
N	14.30%	53.40%
U	4.80%	11.80%
V	1.90%	31.70%
AC	51.20%	32.40%
AD	23.60%	13.60%
AE	11.70%	4.50%
AF	59.60%	77.70%
AG	37.60%	54.10%
AH	15.60%	49.00%
AI	17.10%	16.70%
AJ	38.60%	74.40%
AK	25.40%	70.00%
AL	23.40%	40.60%
AM	5.70%	29.10%
AN	2.10%	47.10%
AO	2.50%	38.90%

SOURCE Authors’ analysis of birth certificate records, Pennsylvania Department of Vital Statistics. NOTES Data confidentiality agreements precluded listing hospitals by name. See Exhibit 1 for more details on hospitals by letter name.

If hospitals are not willing or able to use other service lines to subsidize obstetric units, the pressure to close the units increases. As one informant said, “If you don’t own the hospital and you can’t circle the finances around to support obstetric practice, the obstetric practice is unsustainable.”

ACUTE AND CHRONIC STRESSES RESULTING FROM CLOSURES

Informants identified three major challenges to obstetric units that remained open: increased delivery volume; changes to patient mix and continuity of prenatal care; and issues with staff morale and support (Exhibit 2).

▸INCREASED VOLUME:

Consistent with data shown in Exhibit 1, providers reported “dramatically increased volumes” as the primary challenge from obstetric unit closures. Informants did not distinguish between the closure of the first obstetric unit and the closure of the fourth or the thirteenth unit.

The acute pressures from a closure were exacerbated by two other factors: when the time between the announcement that an obstetric unit would close and the actual closure was just a matter of weeks; and other surges in volume that are part of the normal practice of obstetric care. The increased delivery volumes led informants to report that their obstetric unit “was packed to the gills” and that “it was not uncommon to see patients in a cubbyhole, patients outside my office.” One informant reported that “we had people deliver in the hallway or in the bathroom.” Increases in delivery volume, whether acute or chronic, raised concerns about the quality of care provided to patients.

▸CHANGES IN PATIENT MIX AND CONTINUITY OF PRENATAL CARE:

After the closures, obstetric units that remained open experienced an average increase of 216 percent in women with public or no insurance between 1995 and 2009 (calculated from data in Exhibit 1). Informants confirmed these changes to their overall patient mix, reporting “a shift to more Medicaid patients and just in general a more urban, poor, less insurance, late prenatal care” population. As a result, informants described a greater need for support services, including those provided by social workers.

In addition, after the closures more patients received prenatal care at a different setting from where they delivered their child. In Philadelphia the majority of patients used to receive care at the hospital clinic where they eventually delivered their child.²⁰ After the closures, the system of prenatal and antenatal care was described as “fractured.” One informant said that patients now “get your prenatal care and then you end up somewhere else to have a baby, and it’s not geographically rational.”

As a result, many informants reported that “we are seeing some people with no prenatal care.We are seeing some people who said they had care elsewhere and when we called elsewhere, wherever that may be, they either could not locate their record or told us that there was no care provided there.” Care became more inefficient, which worsened the effects of surges in delivery volume.

▸STAFF MORALE AND SUPPORT:

With increases in delivery volume, new obstacles to patient care, and increased financial pressures, informants reported decreases in staff morale and enthusiasm. Changes to obstetric unit volume resulted in people feeling that their unit was “constantly understaffed” and that it was “not uncommon to do three or four deliveries in a shift…and have a two-patient assignment.”

Additionally, recruiting additional staff to those obstetric units that remained open became difficult. Informants believed that these recruitment issues occurred because of the pressures of seeing patients more quickly with less support, which led to lower job satisfaction and morale.

RESPONSE TO CLOSURES BY REMAINING HOSPITALS

Many responses by hospitals to the challenges mentioned above were reactive rather than proactive. The responses reflect two primary areas of change: structural changes to the obstetric units and changes to the units’ staffing.

▸STRUCTURAL CHANGES:

In response to delivery volume changes, there were a large number of changes to the obstetric units that remained open. However, these changes often occurred months or years after the change in volume at the hospital level, and hospitals “outgrew what they built.” One informant recommended: “Whatever you’re building, build it double what you think you need.You will fill it and regret that your units are not bigger.”

Instead of having its patients recover in labor rooms, one hospital built “a [separate] three-bed recovery room to focus the use of labor rooms for active labor patients [only].” This hospital’s triage area was expanded from a three-bed to a six-bed unit: “We were actually at a point at times that we could use more than six beds. We added an additional ultrasound room because the OB ultrasound volume was going up.”

Another hospital committed extra space and modified it over the past four years. “We’ve gotten support from the hospital, from the administration, and the state to build three recovery rooms.We built an antenatal testing unit [where] we do over 12,000 tests a year, and we built a six-bedded triage unit which we previously hadn’t had,” an informant from this hospital explained.

▸STAFFING MODELS:

Staffing increased on the labor and delivery units. However, respondents reported that they needed a “crystal ball, because we didn’t really know what volume was going to come.” This resulted in significant lag times before staffing could be permanently increased.

For some units, increased hiring was at the level of advanced-practice professionals such as nurse midwives, physician assistants, and nurse practitioners. Other units put new staffing models in place because of the increased volume and concerns about the quality of care and patient safety. “We actually have two obstetrical providers in house 24/7, and three [in house] Monday through Friday, during the day,” one informant explained.

However, the increased number of providers often came at a cost to communication and leadership. “It’s different communicating with eight or nine people than it is [with] four people,” an informant said. “And I think that’s something we’re still working on. You know, how do you get everybody to be part of the big picture? How do you know what’s happening somewhere else and how that’s going to impact your unit or another unit?”

As with the increased delivery volume, there were concerns that these communication challenges could affect the quality of patient care.

Preventing Adverse Effects From Future Closures

Several informants described a severely strained obstetric care system following the widespread closures. Their experience with these closures led them to identify two areas for improvement: better communication between hospitals in advance of closures, and the development of regional solutions to coordinate prenatal care and care at the time of delivery.

COMMUNICATION BETWEEN HOSPITALS TO ANTICIPATE CLOSURES

One key lesson our informants provided was that when an obstetric unit is about to close, leaders from local or state health departments or regional nongovernment organizations need to help improve discussion with fellow institutions and hospital leaders. “If I was closing my maternity program, I would try to reach out to other programs in the area and try to have as smooth a transition as possible,” one informant observed. “It’s a tough decision to close a hospital, and I think hospital administrators would want to work with other hospitals in their region, and also departments should see how we can best care for this population.”

Hospitals compete with each other for obstetric patients. Thus, local and state health departments and other organizations may be needed to facilitate discussions among hospitals to optimize the transition of care for patients in the local community and give remaining hospitals time to adequately staff their units in preparation for potential surges in delivery volume.

REGIONAL WAYS TO FACILITATE EFFICIENT AND HIGH-QUALITY CARE

Informants felt that regional solutions were needed to address the disconnect between where prenatal care is delivered and the hospital where a delivery occurs. One suggested solution was “evidence-based protocols” that would standardize care across providers and institutions. An informant noted that “care is automatically less fragmented if everyone is using the same protocols.”

Regional health information exchange that would facilitate access to prenatal care information and protocols was another solution proposed to decrease the fragmentation of care. Even though such a system of standardized protocols does not exist across hospitals in the region, several informants noted that individual obstetric units should develop “good systems and protocols” to manage both the increased volume and surges in patient numbers experienced by the hospitals that remained open.

Discussion

The case of obstetrical care in Philadelphia County highlights issues that patients and providers may experience when access to care decreases. The primary challenges in Philadelphia included sharp surges in delivery volume, changes in patient mix at individual hospitals, loss of continuity between prenatal and delivery care, and lag times of months to years before new staffing needs could be met and bed capacity could be increased at the hospitals that remained open. All of these challenges contributed to concerns about acute erosion in the quality of care.

Our informants emphasized the importance of communication and planning when even one or two obstetric units close.When a hospital or unit closes without warning, there can be large-scale effects on public health, even if the closure is justified by the hospital’s economic health or the number of deliveries it performed. Communication and planning will become increasingly important with the continued hospital consolidations and mergers that are expected with the ACA’s full implementation.²¹

The greatest challenge identified by our informants was related to delivery volume. Hospitals experienced an average increase in volume of 57.7 percent, and several county hospitals experienced increases of over 70 percent. These volume changes were most disruptive when a closure occurred suddenly and the remaining providers had no warning of it.

The increased number of volume surges reported by our informants parallels similar surges in other areas, such as those caused by emergency department closures in Los Angeles,²² urban hospital closures,²³ and the routine operation of obstetric units²⁴ and provision of pediatric hospital care.²⁵ Similar concerns have been raised when public health departments and hospitals discuss the potential for mass-casualty care.^26,27 Any mechanism that allows hospitals adequate time to adjust their staffing and structural models might help minimize some of the adverse effects of these closures.

Another challenge was that patient mix changed at the hospitals that remained open. They saw an increasing proportion of women having public or no insurance who previously had received care at hospitals that since closed. These changes support informants’ concerns about the financial challenges of running an obstetric unit, which could result in decreased job satisfaction by providers of obstetrical care. Previous surveys of obstetricians suggest that malpractice premiums are a strong driver of job dissatisfaction²⁸ but that job satisfaction also declines as workloads increase and personal control over day-to-day activities at work decreases.²⁹

Potential solutions for these changes include adding specific personnel, such as social workers, to help both patients and providers adapt to these changes. Additionally, hospitals and health care workers need to ensure that they have the cultural competency training to work with changing patient populations.³⁰

We found that the responses from informants at urban academic centers were similar to those from informants at community-based practices both in Philadelphia and in the surrounding suburban counties. This fact emphasizes a need for coordination and communication across county lines and among different hospital systems and public health departments. There is little published literature about such direct discussions between hospitals that otherwise compete for patients within a given health care market.

Obstetrics is particularly challenging because of the time-limited nature of pregnancy. Thus, our informants identified public health departments or other regional collaborative bodies as potential stakeholders in both monitoring changes in access to health care and assisting patients and hospitals in transferring care during pregnancy.

Conclusion

In the future, public health officials should work to identify hospitals and specific units at risk of closing. Furthermore, they should collaborate with affected providers to plan for changes in patient mix and delivery volume to ensure a smooth transition for patients. ■