| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the 1 Departments of Obstetrics and Gynecology, Anesthesiology, and Quality Management, New York Hospital Medical Center of Queens, Flushing, New York.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: We report outcomes before (2000–2001) and after (2002–2005) the introduction of a patient safety program aimed at improving the care of women with major obstetric hemorrhage. Process changes were instituted in late 2001 at the direction of a multidisciplinary patient safety team. A rapid response team was formulated using the cardiac arrest team as a model. Protocols for early diagnosis, assessment, and management of patients at high risk for major obstetric hemorrhage were developed and communicated to staff.
RESULTS: There were significant increases in cesarean births (P < .001), repeat cesarean births (P = .002), and cases of major obstetric hemorrhage (P = .02) between the periods of 2000–2001 and 2002–2005. There was a significant improvement in mortality due to hemorrhage (P = .036), lowest pH (P = .004), and lowest temperature (P < .001) when comparing 2000–2001 with 2002–2005. There were no differences in measures of severity of obstetric hemorrhage between the 2 periods, including Acute Physiology and Chronic Health Evaluation II scores, occurrence of placenta accreta and estimated blood loss.
CONCLUSION: Despite a significant increase in major obstetric hemorrhage cases, we found improved outcomes and fewer maternal deaths after implementing systemic approaches to improve patient safety. Attention to improving the hospital systems necessary for the care of women at risk for major obstetric hemorrhage is important in the effort to decrease maternal mortality from hemorrhage.
LEVEL OF EVIDENCE: II-3
The following cases of maternal deaths occurred in 2000–2001, before our attention to changing the hospital systems. The first patient presented with complete placenta previa, profuse vaginal bleeding, and fetal bradycardia at 28 weeks of gestation and had a maternal heart rate of 110 and blood pressure of 109/70 mm Hg. Hematocrit 2 days before admission was 30% and on admission was 22%. She underwent immediate cesarean delivery under general anesthesia. Intraoperative findings included estimated blood loss of 750 mL, fluid replacement of 2,500 mL, and urine output of 200 mL. She was extubated but remained tachycardic (heart rate 120) and subsequently became oliguric, which did not improve with several fluid boluses. Over the next few hours she became anuric and more severely anemic and developed a coagulopathy. Due to abdominal distension in the setting of the other findings, she was taken back for exploratory laparotomy, with no evidence of intraperitoneal hemorrhage. Despite massive volume and blood replacement subsequently, she became progressively more acidotic until she expired from cardiovascular collapse. The time from admission to death was 21 hours. Her clinical course was compatible with hemorrhage and profound hypovolemia with inadequate initial resuscitation leading to irreversible hemorrhagic shock, coagulopathy, and acute tubular necrosis. Delay in identification of the extent of hypovolemia and vital organ compromise led to inadequate resuscitation.
The second patient was a multigravida at 35 weeks of gestation, with known total placenta previa, a history of previous cesarean delivery, and a high suspicion of accreta, who presented with heavy vaginal bleeding. Upon arrival, her blood pressure was 120/70 mm Hg, and her pulse was 110. The patient had an immediate cesarean delivery. Placenta percreta was obvious and involved the bladder. The patient developed massive intraoperative hemorrhage and became profoundly hypotensive and bradycardic. Lowest intraoperative pH was 6.89, with Pco2 20 mm Hg, Po2 445 mm Hg, base excess –15.6 and hemoglobin 4.2 g/dL. Massive fluid resuscitation occurred with 10,000 mL crystalloid, 500 mL 6% Hespan, and 9 units of packed red blood cells. She required inotropic support with norepinephrine and underwent cesarean hysterectomy. Intraoperatively, the patient developed a coagulopathy treated with 4 units of fresh frozen plasma and 10 units of platelets. Estimated blood loss was 8,000 mL. Postoperatively, the patient remained hemodynamically unstable despite inotropic support and massive transfusion with 1,500 mL crystalloid, 9 units of packed red blood cells, 9 units of fresh frozen plasma, 10 units of platelets, and 100 mL of cryoprecipitate. She had cardiac arrest, and after prolonged attempts at resuscitation, the patient died. The cause of death in this patient was hemorrhagic shock secondary to complications of placenta percreta with multiple organ system failure and coagulopathy. Inadequate initial volume resuscitation and inadequate ongoing volume resuscitation contributed to her death.
These 2 maternal deaths, in the years 2000 and 2001, were due to major obstetric hemorrhage and prompted the creation of a patient safety team that worked to improve the hospital systems for caring for women at risk for major obstetric hemorrhage. We hoped to provide an improved clinical pathway for the identification and management of major obstetric hemorrhage, with the express intent of reducing maternal deaths due to hemorrhage.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Beginning near the end of 2001, a multidisciplinary patient safety team was established that included individuals from the Divisions of Obstetric Anesthesiology, Maternal Fetal Medicine, Neonatology, and the Blood Bank, as well as the Departments of Nursing, Communication, and Administration. A multifaceted approach was used that included the following: 1) We formed an obstetric rapid response team (Team Blue), which was modeled after the cardiac arrest team and included quarterly mock drills on all shifts for various emergency clinical scenarios. 2) We developed clinical pathways, guidelines, and protocols designed to provide for early diagnosis of patients at risk for major obstetric hemorrhage and for streamlined care in emergency situations. 3) In response to a marked increase in the volume of births and gynecologic emergency cases, we separated in-house obstetric and gynecologic responsibilities to allow the in-house obstetrician to focus on obstetric emergencies without fear of possibly neglecting gynecologic emergencies. 4) We formally revised the duties of the in-house obstetrician to include continuous and frequent monitoring of all patients on the Labor and Delivery unit, including those patients who had other private obstetricians. 5) We empowered care providers (including physician assistants, nurses, residents and the in-house attending physician) to immediately involve senior members of the Department whenever there was disagreement with the patient's attending physician's treatment plan (particularly in cases of hemorrhage and possible delay in recognition of the severity of hemorrhage). A senior member of the department then discussed the issue immediately with the attending physician to avoid delay and address problems earlier. 6) Through weekly didactic sessions we educated staff to recognize the stages of hemorrhage described in the Advanced Trauma Life Support Manual6 and disseminated information regarding the new protocols for patient care. The attending, nursing, and ancillary staffs were informed regarding the specifics of the changes, the intent (ie, to improve patient safety), and the importance of early diagnosis of major hemorrhage. 7) We established the role of the Trauma Team. By agreement with the Director of the Trauma Service, the Trauma Team would respond to assist in cases of severe obstetric hemorrhage. This engaged the team that was the most experienced in resuscitation of patients with hemorrhagic shock within our institution. The team included junior and senior surgical house officers under the direction of the surgical trauma attending physician. Members of the Trauma Team are expert in the placement of large-bore intravenous lines (by venous cutdown if necessary), are knowledgeable about the physiology of volume resuscitation, and assist in acquiring adequate amounts of blood products for massive blood replacement.
The protocols and guidelines created included the following: 1) We prepared for major hemorrhage in patients with known placenta previa (Fig. 1). This preparation included antenatal consultation with Maternal Fetal Medicine, senior gynecologic surgeons, and Obstetric Anesthesiology; ultrasonography to identify placenta accreta in patients with prior uterine surgery; twice-weekly type and screen to allow for more rapid availability of blood products in the event of major hemorrhage; and planned cesarean delivery at 36 weeks after amniocentesis for fetal lung maturity. 2) We prepared for major hemorrhage in patients with suspected placenta accreta (Fig. 1). This included weekly autologous blood donation; erythropoietin, iron, and vitamin therapy; possible consultation with interventional radiology, judicious placement of extra intravenous lines, and a 7.5 French internal jugular cordis for invasive monitoring and volume replacement; intraoperative monitoring with an arterial line and central venous pressure; and transfer to the postanesthesia care unit or surgical intensive care unit as needed. In addition, we employed the Cell Saver (Haemonetics Corporation, Braintree, MA), but this was used only after delivery of the fetus and after copious peritoneal irrigation had been performed.3 The use of weekly autologous blood donation was to prevent bloodborne infection and contributes to resolving any potential shortage of blood in our area. 3) We obtained peripartum or intraoperative consultation with the Trauma Team as necessary. 4) We counseled patients with suspected placenta accreta about the likely decreased maternal mortality of planned cesarean hysterectomy.7 For patients who agreed, planned cesarean hysterectomy was then performed. 5) We scheduled cesarean delivery and cesarean hysterectomy in the main operating room under the direction of senior gynecologic surgeons (Fig. 1). Staff and facilities of the main operating room are better equipped to assist in performing hysterectomy than the Labor & Delivery suite, and this avoided the problem of consuming staff and resources on Labor & Delivery necessary for the care of other patients.
|
Data on obstetric volume, mode of delivery, major obstetric hemorrhage, and other outcomes were collected from 2000–2005. Cases were identified prospectively through an ongoing Quality Assurance program for the entire patient cohort (2000–2005), but demographic and outcome data were recorded from medical records during the period of January, 2000 to May of 2001. Demographic and outcome data were recorded prospectively on each patient beginning in June 2001. The ongoing quality assurance program involves daily monitoring by senior departmental leaders. Senior departmental leaders receive reports on a daily basis from care providers regarding all cases. Cases of major obstetric hemorrhage were highlighted and these cases were included in the database as they occurred. Outcomes analyzed included maternal deaths, lowest documented maternal pH, lowest documented maternal temperature, and the occurrence of coagulopathy.
Our definition of major obstetric hemorrhage for the purposes of this report included 1 or more of the following: estimated blood loss of 1,500 mL or more, need for blood transfusion, need for uterine packing, performance of uterine artery ligation, and performance of cesarean hysterectomy. This definition is different from that of postpartum hemorrhage, so that the rate of major obstetric hemorrhage by our definition is expected to be lower than the known incidence of postpartum hemorrhage. This article describes a review of cases before and after intervention planned as part of patient safety activities and thus the Institutional Review Board deemed that approval was not necessary. Data were analyzed by Fisher exact test, Wilcoxon rank sum test for comparing 2 unmatched samples, and unpaired t test as appropriate. Data were compared between the 2 epochs before and after the systemic changes were implemented, 2000–2001 compared with 2002–2005. Analysis was performed at an interactive statistical calculation Web page: http://members.aol.com/johnp71.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Clinical characteristics, measures of severity of hemorrhage, and outcomes are shown in Table 2. The patient groups from the 2 periods (2000–2001 and 2002–2005) were similar as measured by age, parity, and incidence of prior cesarean delivery. The severity of obstetric hemorrhage also seemed to be similar between the periods, as measured by Acute Physiology and Chronic Health Evaluation II scores,8 occurrence of placenta accreta, and amount of estimated blood loss (Table 2).
|
Maternal deaths were significantly reduced in the period after the systemic change (P = .036). In addition, there were significant differences in lowest pH (P = .004) and lowest temperature (P < .001). There were 3 maternal deaths during the entire study period, 2 due to hemorrhage detailed above, and 1 due to thromboembolism in a patient who had sudden cardiovascular collapse on postpartum day 1 after cesarean delivery. Aggressive resuscitation was attempted, but unsuccessful. Medical Examiner's report demonstrated a massive pulmonary embolus.
The 2 periods were also analyzed according to operative characteristics, such as need for cesarean hysterectomy, volume of transfusion, operative time, need for intubation for greater than 24 hours, and number of hours intubated (Table 2). There were no significant differences in these measures between the periods of 2000–2001 and 2002–2005. Throughout the entire study period (2000–2005) the rate of peripartum hysterectomy was 1.3 in 1,000 (24/18,723). Placenta accreta with prior cesarean delivery accounted for 14 of 24 (58.3%) cases of cesarean hysterectomy. We suspected accreta in 7 cases and confirmed it in 4 cases at delivery during the entire study period 2000–2005. This level of accuracy in assessment for possible accreta is consistent with published literature. The operative characteristics, morbidity, and mortality of patients undergoing peripartum hysterectomy are shown in Table 3. There was a significant difference in lowest pH in patients undergoing cesarean hysterectomy between the periods 2000–2001 and 2002–2005.
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our findings indicate that there were significant improvements in outcomes in the period after introduction of systemic changes, including improvements in maternal deaths, lowest pH, and lowest temperature. This occurred despite finding no difference in measures of severity of obstetric hemorrhage and significant increases in the number of cases of major obstetric hemorrhage between the study periods. When comparing the time periods before and after the systemic changes, there were significant differences in lowest temperature and in lowest pH (Table 2), and we believe this suggests that the team's response to massive hemorrhage improved after the interventions we describe. All of the above findings were statistically significant, but the reduction in maternal mortality is not a robust observation, because there were only 2 deaths.
The process of implementing the systemic changes was time intensive and required considerable effort by many individuals. The patient safety team met numerous times and deliberated on the specific nature of our response. The efforts also included repeated education of all care providers on the diagnosis and management of hypovolemic shock. These educational efforts also involved a substantial amount of time, thought, and teaching.
Although an unavoidable consequence of the retrospective nature of the demographic and outcomes data collection in 2000 and 2001 is that there may be ascertainment bias in our reporting of results, this bias does not apply to the most important outcome of maternal deaths, and hemorrhage case identification was contemporaneous throughout the study. This study design analyzes outcomes before and after changes in management and does not allow a determination of which of several interventions may have accounted for improvements in outcome. We feel the current data support the conclusion that a well-reasoned, carefully constructed, and multifaceted program focusing on patient safety can improve outcomes, although we cannot attribute any specific improvement to any specific change that we instituted.
Although the risk of placenta previa with or without accreta in patients with multiple cesarean deliveries is difficult to quantitate,10 recently published prospective data11,12 corroborated previously published retrospective data on the substantial risk of accreta associated with previa and prior cesarean.13 Placenta previa is a detectable condition, allowing for a preventive clinical pathway such as that developed in Figure 1 to be implemented.
In August, 2004, the State of New York Department of Health and The New York City Department of Health and Mental Hygiene issued a memorandum about the prevention of maternal deaths due to hemorrhage.5 This memorandum details the necessary hospital systems that may provide a safety net for maternal patients. Our institutional changes preceded this memorandum by 3 years and included each of the suggested changes of this memorandum plus additional measures.
Principles of quality improvement require that "systems" thinking take place when confronted with adverse outcomes. It is tempting to correct the proximate cause without addressing systemic opportunities to improve. We believe that these data support a systemic response and hope they are useful to others faced with the task of improving safety in obstetric suites.
|
Footnotes |
|---|
Poster presentation at the annual meeting of the Society for Maternal–Fetal Medicine, Reno, Nevada, February 10–12, 2005.
Corresponding author: Daniel W. Skupski, MD, Department of Obstetrics and Gynecology, the New York Hospital Medical Center of Queens, 56-45 Main Street, Room M-372, Flushing, NY 11355; e-mail: dwskupsk{at}med.cornell.edu.
doi:10.1097/01.AOG.0000215561.68257.c5
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Miller DA, Chollet JA, Goodwin TM. Clinical risk factors for placenta previa-placenta accreta. Am J Obstet Gynecol 1997;177:210–4.[Medline]
3. ACOG Committee on Obstetric Practice. ACOG Committee Opinion. Number 266, January 2002: placenta accreta. Obstet Gynecol 2002;99:169–70.[Medline]
4. Forna F, Miles AM, Jamieson DJ. Emergency peripartum hysterectomy: a comparison of cesarean and postpartum hysterectomy. Am J Obstet Gynecol 2004;190:1440–4.[Medline]
5. Frieden TR, Novello AC, King J. Health Alert: Prevention of maternal deaths through improved management of hemorrhage. Letter from State of New York Department of Health and The New York City Department of Health and Mental Hygiene, August 9, 2004.
6. American College of Surgeons. Committee on Trauma. Advanced trauma life support for doctors: ATLS. 6th ed. Chicago (IL): American College of Surgeons; 1997.
7. Sheiner E, Levy A, Katz M, Mazor M. Identifying risk factors for peripartum cesarean hysterectomy: a population-based study. J Reprod Med 2003;48:622–6.[Medline]
8. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818–29.[Medline]
9. Kohn LT, Corrigan JM, Donaldson MS. To err is human: building a safer health system. Washington (DC): National Academy Press; 1999.
10. Greene MF. Vaginal birth after cesarean revisited. N Engl J Med 2004;351:2647–9.
11. Silver RM. The MFMU cesarean section registry: maternal morbidity associated with multiple repeat cesarean delivery. Abstract. Am J Obstet Gynecol 2004:191:S17.
12. Rashid M, Rashid RS. Higher order repeat caesarean sections: how safe are five or more? BJOG 2004;111:1090–4.[Medline]
13. Clark SL, Koonings PP, Phelan JP. Placenta previa/accreta and prior cesarean section. Obstet Gynecol 1985;66:89–92.
This article has been cited by other articles:
|
|
 |
|
  J. M. O'Brien Placenta Previa, Placenta Accreta, and Vasa Previa Obstet. Gynecol., January 1, 2007; 109(1): 203 - 204. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Includes the counseling below and a recommendation for cesarean hysterectomy. Low parity may decrease the strength of the recommendation if future child bearing is desired. 
Includes bed rest, pelvic rest, preparation for cesarean delivery, serial complete blood count, consider erythropoietin, iron, and vitamin supplements, and serial autologous blood donation.