11 Dec

What Have We Learned & Where Are We Going?

by Emily Hamilton, MDCM
Senior Vice President, Clinical Research, PeriGen

“In the New Year, you carry all the experiences of the past years and that is the greatest power of every New Year! This year, you are less student and more master!”
~Mehmet Murat Ildan, Contemporary Novelist & Playwright


Flling rates of intrapartum-related neonatal encephalopathy

Figure 1. Falling rates of intrapartum-related neonatal encephalopathy

Just as individuals gain experience and wisdom on life’s journey, so do organizations, professions and companies. Experience helps because as professionals we are entrusted with responsibility both at a personal and organizational level to provide safe maternity care. Among the myriad of challenges crying for our attention, which ones are critical, have a chance for success, or bring the greatest benefit for our mothers, babies and their families? With the New Year beginning it is time to pause and reflect on some of the lessons learned and past accomplishments because they will help us plan our journey for the coming year. As you read on, note the clarity of insights gained by using “Big Data” and “Wide Lenses” for time and geography.

One of the United Nation’s Millennium Goals – to reduce child mortality, has brought us more reliable data on an important obstetrical topic: namely, intrapartum-related neonatal encephalopathy (IP-NE). On a world scale, intrapartum-related hypoxic events are estimated to have a huge impact – contributing to ¼ of neonatal deaths and ½ of late pregnancy stillbirths. They make the largest single-condition contribution to disability adjusted life-years. 1-4

When one examines progress over 20 years in multiple countries, steady incremental improvement is evident. Figure 1 shows the falling rates of IP-NE in regions including and most similar to the US1 Rates in all the other geographical regions were much higher. They all showed progressive descent fell except for the region of Sub-Saharan Africa.

Two other recent and large studies provide further important insights regarding the nature of IP-NE and related clinical care. 5, 6 Researchers in Sweden, where infant mortality rates are generally amongst the best in the world, examined 71,189 birth records to investigate the association between EFM patterns and neonatal outcome. 5 The rates of neonatal encephalopathy were very low at 1.1 per 1000 births. They concluded that moderate and severe encephalopathy was attributable to asphyxia in 60% of their NE cases and most evolved during labor.

Moving closer to home, Clark et al examined the effect of compliance vs non-compliance with an oxytocin administration protocol in 14,398 women undergoing induction of labor at HCA.6 Oxytocin misuse is a common finding in births with severe metabolic acidosis. 7-9 Furthermore, oxytocin misuse is a modifiable risk factor. In this most recent study compliance was associated with:

  • Fewer admissions to the NICU (3.8% vs  5.2% P=0.01)
  • Fewer low Apgar scores (4.9% vs 6.4% P=0.04)
  • Fewer cesareans     (15.8% vs 18.8% P<0.01)
Figure 2. PeriCALM Patterns display showing the analysis related to uterine tachysystole.

Figure 2. PeriCALM Patterns display showing the analysis related to uterine tachysystole

Members of our own PeriGen family have experienced similar success stories. 10 MedStar Franklin Square Medical Center launched a highly successful IT initiative focusing on uterine tachysystole (overly frequent contractions). PeriCALM Patterns fetal monitoring software with built-in pattern recognition and specialized long-term graphical displays was introduced in 2011. At a glance any clinician could detect if uterine tachysystole (UT) was present, if it was transient or persistent, if it provoked fetal heart rate decelerations and see the Montevideo units as is shown in Figure 2.

This study involved the systematic re-examination of each 30-minute segment of tracing from all 10,518 monitored term labors. Comparing years before immediately before and after the introduction of PeriCALM Patterns they observed that:

  • The rate of UT with oxytocin fell from 22.7% to 17.3% P<0.0001.
  • Average duration of UT fell from 64 minutes to 54 minutes.
  • Total time spent in UT fell by 36.5%.

Long term trends analysis is a key component of helping humans see significant trends and to anticipate what is about to happen. In psychological parlance this ability to assess current conditions and project the most likely development is known as situational awareness. In medical behavioral studies it is one of the key skills required across all acute care settings.

Computers are masters of rapid computation and efficient data visualization. Anecdotes from nursing staff related how the displays made easy it was to see when uterine tachysystole was present or even about to occur. Continuous calculations removed the subjectivity of choosing where to count the contraction rates. Quantitative data facilitated communication because predominant patterns were clearly evident. Nurses had the authority to stop oxytocin when uterine tachysystole occurred.

There is abundant evidence today that the human brain has a fixed functional capacity and that every mental task we perform detracts from our ability to do another at the same time. 11 Computers are very consistent at number crunching, that is counting, adding, measuring or doing exactly what they are programmed to do. In contrast, humans are very good at reasoning based on deep clinical understanding Clinicians will always have to integrate EFM tracing findings with other clinical information. That said, the repetitive analytic capacity of PeriCALM Patterns and its long-term displays are useful adjuncts for clinicians. This study demonstrated the synergistic effects of this technology and dedicated clinicians. Together they achieved an impressive reduction in UT.

In summary, what lessons have we learned?

Lesson 1. It is impossible to see progress in rare conditions using small samples over short periods of time.  Rates derived from large data sets and trends over long periods are much more revealing.

Lesson #2. A substantial portion of NE arises and evolves during labor.

Lesson #3.  The incidence of intrapartum-related NE, a devastatingly serious condition, is falling.

Lesson #4.  Clinical actions matter. Oxytocin protocol compliance is associated with better neonatal condition and lower cesarean rates. Uterine tachysystole rates can be reduced.

Lesson #5.  Computers and clinicians can be highly synergistic.

No one is suggesting that technology replace clinicians. However, computerization is useful for tasks like calculations, data visualization, reminders and communication thus freeing clinicians to focus applies their energy on higher level reasoning and clinical interventions.

Maternal child health care has come a very long way. Detractors of EFM often refer to randomized clinical trials conducted more than 30 ago that showed no neonatal survival benefit with EFM, although EFM use was associated with a substantial reduction in the incidence of neonatal seizures. In the largest of these studies, deaths or seizures occurred at an astonishing rate of 1 in 225.12 Today, intrapartum fetal death is exceedingly rare and moderate and severe IP-NE are estimated to occur in approximately 1.5/1000 births among the High-Income countries and 0.6/1000 specifically in the Swedish study. This achievement did not happen by chance. Concerted efforts on many fronts have contributed. It happened in part, because most clinicians did not adopt a defeatist attitude such as believing that the origins of IP-NE were beyond our influence or that EFM held no value. Rather these clinicians continued to find better ways to fine tune the imperfect tools that are available and improve our health care systems.

Healthcare informatics has also evolved. We have seen a maturation of software for basic hospital wide electronic medical records and convergence on a few products. With this accomplishment we have seen an increasing demand for the efficient overlay of smart modules with real clinical benefits. In parallel, the PeriGen experience has confirmed our belief that the most useful smart software applications are those that harness the power of computers to do what they do best especially in areas that humans find challenging or time consuming.

As for the New Year, we will continue to apply sound analytical methods on high impact clinical issues where objective quantitative analysis helps clinicians see critical factors or developing trends and intervene in a timely fashion. We have reached a new level in medical informatics and a very exciting one indeed. Stay tuned as we work to bring these ideas to the bedside.


1. Lee AC, Kozuki N, Blencowe H, Vos T, Bahalim A, Darmstadt GL, Niermeyer S, Ellis M, Robertson NJ, Cousens S, Lawn JE. Intrapartum-related neonatal encephalopathy incidence and impairment at regional and global levels for 2010 with trends from 1990. Pediatr Res. 2013 Dec;74 Suppl 1:50-72. doi: 10.1038/pr.2013.206.

2. Wn J, Shibuya K, Stein C. No cry at birth: global estimates of intrapartum stillbirths and intrapartum-related neonatal deaths. Bull World Health Organ. 2005 Jun;83(6):409-17. Epub 2005 Jun 17.

3. Lawn JE, Lee AC, Kinney M, Sibley L, Carlo WA, Paul VK, Pattinson R, Darmstadt GL. Two million intrapartum-related stillbirths and neonatal deaths: where, why, and what can be done? Int J Gynaecol Obstet. 2009 Oct;107 Suppl 1:S5-18, S19. doi: 10.1016/j.ijgo.2009.07.016

4. Murray CJ, Lopez AD. Global mortality, disability, and the contribution of risk factors: Global Burden of Disease Study. Lancet. 1997 May 17;349(9063):1436-42.

5. Jonsson M, Agren J, Nordén-Lindeberg S, Ohlin A, Hanson U. Neonatal encephalopathy and the association to asphyxia in labor. Am J Obstet Gynecol. 2014 Dec;211(6):667.e1-8. doi: 10.1016/j.ajog.2014.06.027. Epub 2014 Jun 17.

6. Clark SL, Meyers JA, Frye DK, Garthwaite T, Lee AJ, Perlin JB. Recognition and response to electronic fetal heart rate patterns – impact on newborn outcomes and primary cesarean delivery rate in women undergoing induction of labor. Am J Obstet Gynecol. 2014 Nov 22. pii: S0002-9378(14)02249-2. doi: 10.1016/j.ajog.2014.11.019. [Epub ahead of print]

7. Clark SL, Belfort MA, Dildy GA, Meyers JA. Reducing obstetric litigation through alterations in practice patterns.  Obstet Gynecol. 2008 Dec;112(6):1279-1283.
8. Berglund S, Grunewald C, Pettersson H, Cnattingius S. Severe asphyxia due to delivery-related malpractice in Sweden 1990-2005. BJOG. 2008 Feb;115(3):316-323.

Jonsson M, Nordén SL, Hanson U. Analysis of malpractice claims with a focus on oxytocin use in labour. Acta Obstet Gynecol Scand. 2007;86(3):315-319.

10. Smith S, Bunting K, Hamilton E. Using Intelligent Electronic Fetal Monitoring Software to Reduce Iatrogenic Complications of Childbirth: A Case Study.J Healthc Inf Manag, in press

10. Rock D., Your Brain at Work: Strategies for Overcoming Distraction, Regaining Focus, and Working Smarter All Day Long Harper Business. Harper Collins, New York, NY.

12. MacDonald D, Grant A, Sheridan-Pereira M, Boylan P, Chalmers I. The Dublin randomized controlled trial of intrapartum fetal heart rate monitoring. Am J Obstet Gynecol. 1985 Jul 1;152(5):524-39

24 Jun

Obstetrics: overlooked ACO domain

Obstetrics is one area in which ACOs can make a huge difference.

ACO approach promises to reduce risks, lower cost

June 23, 2014

There are many dimensions to the Accountable Care Organization challenge, including the logistics of changing a well-established acute care model, the process of configuring the network, analyzing IT capabilities across the spectrum and determining how all those moving parts will work together.

It’s a tall order indeed, which is why ACOs remain largely in a gestational state.

Another equally important part of the formation process is in risk assessment and how an ACO will manage a value-based performance model in a way that it generates the same returns as fee-for-service, said Matthew Sappern, CEO of Cranbury, N.J.-based PeriGen.

“Healthcare executives are still trying to get their arms around the definitions of ACOs,” he said. “My definition is that it involves risk management — how are you managing risk no that pay-for-performance is front and center. Risk mitigation is the cutting-edge issue here.”

Population health and its big data subset are of primary concern to the provider community, and Sappern acknowledges that the attention being paid to it is warranted. Even so, he says there is one area that is being vastly overlooked in its potential to dramatically reduce risk and save costs: Obstetrics.

“OB accounts for the most of the malpractice lawsuits, and legacy systems have perpetuated the risk exposure,” he said. “It does nothing to help clinicians understand what is going on and doesn’t help nurses spot emerging patient trends. That adds up to more exposure and a negative effect on the balance sheet.”

The OB field is unlike any other in medicine, which makes it a clinical outlier in ACO equations. The patients are primarily young, giving birth is a natural body function rather than an illness and labor status can be a critical minute-by-minute monitoring process. Real-time accurate information is paramount and ironically, is not often available to OB nurses, Sappern said.

“OB nurses have tough jobs,” he said. “They have multiple patients and they are trying to keep track of key data elements that are changing all the time. They need to spot non-reassuring trends and do something about them immediately — OB cases can go wrong quickly.

To eliminate the guesswork and “hyper-vigilance” associated with OB cases, PeriGen has developed a solution that applies computing algorithms to interpretations of maternity ward patient developments. This way nurses can focus on their duties without constantly checking on how different patients are doing, Sappern said.

“We allow the computer to count the factors going on and steer the nurse toward potential concerns,” he said.

To be sure, system disparity still exists within enterprises and across the ACO ecosystem, agrees Nalin Jain, delivery director of advisory services for Buffalo, N.Y.-based CTG Health Solutions.

“Siloes and disparate systems remain a challenge, but those are technical details to be ironed out after successfully moving toward a performance-based culture and attitude,” he said. “You need to first get the right governing structure and get people doing the right things. That is the work that needs to be done first.”

To appreciate the magnitude of difficulty involved with forming ACOs, Jain refers to the 1980s and ‘90s, when acute care entities experimented with acquiring post-acute properties, but soon gave up because the methodologies and reimbursement structures were too different. Now, however, Jain believes these organizations have a chance at success because technology has forged better connections and fostered greater understanding about what is needed to make them work.

“If accountable care is a verb and not a noun, then you can see it is just good medicine,” he said. “In advising our clients on what is needed for ACOs, we are very clear about what lies ahead and that they should have a triple aim of maintaining a caring population and providing a positive patient experience while cutting healthcare costs.”

Nick Stepro, director of analytics at Burlington, Mass.-based Arcadia Healthcare Solutions, also sees a massive challenge for providers in aggregating data from disparate, siloed entities within the ACO confines.

“Over the past year, health systems engaged in ACOs or similar care models have begun to understand the real challenge of getting timely visibility into the navigation, costs, and health status of their populations,” he said. “In many cases, this means integrating data from thousands of providers across of EHRs, hospital systems, and claims feeds. ACOs that had not invested in these type of capabilities were effectively flying blind, with no visibility into performance until it was time to submit data at year-end.”

In order to avert getting this unpleasant surprise, the health systems involved in ACOs are now investing in a clinical data integration strategy to enable real-time performance monitoring and care coordination, Stepro said.

As experts point out, aging infrastructure prevents the interoperability facet necessary for facilitating ACO communications and coordination – in essence, the healthcare industry is being trapped by its past and the technology of yesterday is preventing forward progress. Another major inhibitor is the prevention of seamless workflow, says John Gobron, CEO of Denver-based Aventura.

“It seems like everyone benefits from the digitization of data except clinicians, who have found it to be a cumbersome process,” he said. “The computer shouldn’t be an impediment.”

There is still too much complexity in workflow, Gobron says, with too many levels and layers between users and the right information. That is why he says Aventura is focused on improving workflow so that clinicians can get the data they need in as easy a manner as possible.

“The problem is easier to solve when it concerns the data instead of the infrastructure,” Gobron said.

“But the workflow and usability factors just are not there yet. We’re still trying to shake off the paper age. Digitization doesn’t mean electronically generated paper. This is more than just data entry. Until the workflow issue can be solved, it just won’t happen.”


30 May

Reducing Birth Brain Injuries with EFM Pattern Monitoring


Patterns 2

By Emily Hamilton, MDCM, Senior Vice President of Clinical Research, PeriGen

Birth-related brain injury, although rare, carries a high rate of devastating consequences for all involved. These consequences can include lifelong disability for the child, family breakdown, multi-million dollar settlements in litigation, and a demoralized medical professional who oftentimes elects to cease obstetrical practice.

An astounding and sad fact is that roughly half of these cases are potentially preventable. The leading issue is a failure to correctly assess the baby’s tolerance to labor based on continuous recordings of fetal heart rate and maternal contractions.

Commercial electronic fetal monitors (EFMs) that produce the continuous recordings of fetal heart rates and contractions were introduced in 1968; today, they are used in almost all births. Hence, how is it possible that a technique in place for nearly half a century is still misused to this degree today? If the answer were simple the problem would have been solved long ago.

The Underlying Problem

Problems with determining when an intervention is required fall into four categories:

  • Fetal InaccessibilityClinicians cannot examine the state of baby’s brain in labor, nor measure the amount of oxygen being delivered to it. They can only measure what is accessible — the fetal heart rate — which is a poor substitute for the actual information they need.
  • Inconsistent Visual Analysis. For decades clinicians have examined miles of paper recordings to identify what aspects or features of the tracings are associated with poor outcomes. Thousands of publications ensued. Although there is general agreement on some basic correlations, this methodology stymies any significant advance in reproducible results. Studies will always be limited by the imprecision and inconsistency of inspection with the naked eye as well as confined to relatively small study sizes.
  • Impractical Clinical Guidelines. Clinical guidelines on how to classify and manage tracings leave much to be desired. It is impractical and unrealistic to expect clinicians at the bedside to apply complex algorithms based on dozens of rules. A simple three-level system is more manageable and commonly used. Unfortunately, the great majority of tracings reside in the middle level, where abnormalities range from minor to highly abnormal. This simple classification method does not help clinicians chose clinical management because minor midlevel abnormalities require little or no intervention, whereas highly abnormal midlevel changes require urgent intervention. Thus, choosing clinical management for tracing in the “middle” category is akin to deciding what to wear given a nebulous weather forecast of “rain,” which could mean anything from light showers to a hurricane.
  • Poor Projection.  In most diagnostic situations the clinician’s task is to determine whether a condition is present. Correct usage of EFM is far more complex because the objective is to prevent a condition, not wait until diagnosis is certain. That is, clinicians must project what could happen to the baby if the tracing continues or deteriorates. If the baby’s tracing deteriorates, clinicians must determine the length of time the labor could take and at what point before delivery the baby should be removed from the stresses of labor.

Expecting clinicians to make these projections at many points during each labor — and frequently in the middle of the night when human judgment — is most vulnerable to fatigue, and based on very nonspecific data with generally unhelpful guidelines, creates a “perfect storm” for human error.

Resolving the Perfect Storm

Two decades ago, researchers at McGill University, the leading Medical and Doctoral University in Canada, began a collaboration to address these issues. At that time the university research leaders were highly supportive of developing multidisciplinary teams to find solutions to high impact clinical problems.  In studying obstetrical practice challenges, it was clear automated EFM pattern recognition was an essential step to address the issue of human inconsistency and imprecision in tracing analysis.  Previously, many groups had tried to develop such software but the exercise had proven difficult to achieve adequate accuracy.

Their top-notch team of mathematicians, engineers, computer scientists and clinicians, aligned by a common mission and abundant persistence, created EFM pattern recognition software that achieved a level of accuracy far exceeding OB reported research findings. This EFM pattern recognition software is embedded today in Cranbury, N.J.-based PeriGen’s fetal surveillance and archiving software PeriCALM®  Tracings™. The solution is in clinical use in more than 200 hospitals’ labor and delivery (L&D) departments across the U.S.  PeriCALM Patterns is the only fetal strip interpretation technology validated independently and favorably by the National Institutes of Child Health and Development (NICHD). The NICHD study reviewed the analysis of PeriCALM Patterns during the final hour of 100 tracings, and concluded that, “computerized fetal heart rate interpretation has substantial agreement with experts’ evaluation.”

Seeing Results in Real Time

In addition to enhancing clinical efficiency through real-time analysis at the bedside, PeriGen hospital clients have begun to realize tangible benefits such as a substantial drop in uterine tachysystole rates.  Uterine tachysystole refers to overly frequent contractions, often caused by the medication oxytocin.  Oxytocin, an effective drug to increase the rate and strength of uterine contractions, is used in more than half of all labors. Despite well-established guidelines, incautious use of oxytocin is reported in 45 to 71 percent of births with severe asphyxia and subsequent litigation. Failure to comply with standard oxytocin guidelines generally makes these cases legally indefensible and hence extremely costly.

PeriGen’s PeriCALM Patterns attacks this problem directly via specialized analysis, displays and visual alerts. Clinicians can instantly see onscreen when uterine tachysystole occurs, evaluate how the baby is responding to it, and then see the immediate effects of modified oxytocin dosages.

Brighter Future for Mothers and Babies

OB Research Enters Big Data Era

Obstetrics medicine has entered today’s “big data” era in which hospitals are amassing vast amounts of electronic patient record data including digital EFM tracings. These massive data sets are essential when studying rare events with multiple contributing factors. Yet they are only part of the solution. An automated method of analysis is also required, because these large data sets cannot be analyzed consistently or precisely with traditional visual inspection.

PeriCALM’s EFM capability has driven a resurgence of research producing a better understanding of EFM characteristics reliably associated with severe neonatal depression or metabolic acidosis. Automated detection of EFM features and the identification of clusters or trends that are truly predictive of poor outcomes would go a long way toward addressing the human inconsistency in bedside care and the devastating consequences it can incur.

The original dream of conducting large-scale studies has materialized. Within the last decade, PeriCALM Patterns and related software have spawned more than two dozen peer reviewed articles. The PeriGen OB solutions are central components in ongoing large studies involving several leading academic medical centers.

Potential clinical impact is also easier to assess. Clinical trials of the size required for rare-outcome research are prohibitively expensive. Retrospective analysis of “big data” is now a realistic alternative to costly classical prospective clinical trials, and in turn can hasten the time to bring research findings to everyday clinical care.

In many ways the historical situation created the conditions of a “perfect storm.” That storm persisted for a very long time because we lacked the tools and data to find better ways. A brighter day is dawning as we now have multiple sophisticated OB tools to counter the obstacles that created the impasse of the past decades as well as tangible evidence of positive results.

Moreover, modern information technology, cloud-based computing and mobile devices can bring these discoveries to assist clinicians in real time at the bedside. Everyone can benefit from the advances gained from rigorous analysis of “big data” now fueling exponential innovation. Stay tuned for more game-changing developments in obstetrical informatics from PeriGen.

Dr. Emily Hamilton directs PeriGen research teams to develop and refine innovative decision support-based fetal monitor technologies designed to prevent birth-related injuries such as shoulder dystocia. Under Dr. Hamilton’s leadership and vision, risk-reduction perinatal technology systems have come of age in obstetrics care practice.

Read similar articles at http://www.rxeconsult.com/

15 Jan

Healthcare IT and the Promise and Curse of Dimensionality

HIMSS - Transforming Healthcare Through IT

January 13, 2014

by Emily Hamilton, MDCM, FRCSE, FACOG, Senior Vice President of Clinical Research, PeriGen

Assessing the relationship between a suspected agent and a particular disease is a fundamental part of medicine. Treating causes of disease is beneficial. When a relationship is merely one of association, accordingly there is no therapeutic potential related to that agent. In essence, coloring gray hair does nothing to cure dementia.

Controversy over the role of Vitamin D exemplifies this distinction. Low blood levels of Vitamin D have been associated with poorer outcomes in hundreds of observational studies. In turn, Vitamin D has been promoted as a therapeutic agent for many conditions, including osteoporosis cancer, cardiovascular disease, diabetes and dementia, to name a few. Still, other studies report supplementation with Vitamin D did not result in improved health, concluding that low levels are merely a consequence, not a cause, of poor health.

While the jury is still out on the Vitamin D debate, it does demonstrate the complexity in understanding what actually determines a health outcome and the hazards of a simple single-factor model of disease.

The outcomes of chronic ailments have multiple influencing factors grouped as patient related (e.g., age, gender, family history, stage of illness, etc.), nature of care (i.e., was the care administered correctly, on time, by procedure and skillfully?), and interfering (e.g., comorbidities, medications, toxic substances, lifestyle, etc.). Each factor is a statistical dimension contributing to a complex picture of the disease process.  However, to sort out the effect of each dimension and the effect of combinations of factors require vast amounts of data.  The curse of dimensionality is that the amount of data needed to support the analysis often grows exponentially with the number of dimensions.

The value of healthcare information technology

As a result of extensive aggregation of data encompassing the three groups of influencing factors, clinicians now more fully understand causes of health outcomes in ways that were never before possible. Big data, computational capacity and modern statistical techniques pack a powerful punch. Obstetrics is well suited to reap the benefits of this trio because it focuses on one major condition, childbirth, which evolves over a matter of hours, not decades like chronic diseases.

Remarkable examples of this trio at work include the analysis of factors associated with births where the baby showed signs of poor tolerance to labor. In a large study of more than 100,000 births, several factors were found to have an independent effect; moreover, many of them tended to cluster at night. Awareness of this toxic combination can help OB staff better anticipate these potential problems and proactively address the underlying factors under their control. Figure 1(below) illustrates the rate of neonatal depression over two consecutive days. Note the strong circadian patterns reflecting the interaction of factors clustering at night.

Another example of healthcare IT at work is electronic fetal heart rate monitoring used by clinicians to assess the baby’s tolerance to labor. Until recently, clinicians relied upon scanning paper tracings with the naked eye to spot worrisome trends. Not only is this method notoriously inconsistent, it made large analytical studies virtually impossible to do. To provide perspective, consider this: Fetal heart rate tracings from a mere 5,000 births can easily produce over 20 miles of paper strips and contain well over 3 million different defined features. It is no wonder that fetal heart rate monitoring research had stalled until the advent of computerized pattern recognition and computational methods designed to tackle this analytical problem.

The value of healthcare IT reaches far beyond discovery. Frontline clinicians caring for patients at the bedside can benefit from real-time prospective decision support based on new discoveries as well as a host of other intelligent reminders based on best practices and internal computations. In addition, OB care teams or their managers can use these systems to measure trends in performance and outcomes and see the impact of their interventions.

Call to Action in 2014

Thanks to advanced information technologies, the collection and organization of vast amounts of data is a reality today. Extensive data sets combat the curse of dimensionality and facilitate its promise of a better understanding of complex issues.  Physicians, nurses, care teams and IT professionals all have a role to accelerate this process.  Clinicians need to develop and use a common language so that data can be aggregated correctly. Although the task of entering correct and complete data rests with the clinicians, the IT teams must support them by designing systems that ensure usability. Good data is golden and access to it helps all healthcare stakeholders – and most importantly, our patients.

Rate of Neonatal Depression versus Time of Birth c

About the Contributor

Dr. Emily Hamilton directs PeriGen research teams to develop and refine innovative decision support-based fetal monitor technologies designed to prevent birth-related injuries such as shoulder dystocia. Under Dr. Hamilton’s leadership and vision, risk-reduction perinatal technology systems have come of age in obstetrics care practice.

16 Sep

Reducing Elective Early-Term Deliveries: One Hospital’s Groundbreaking Optimization Work


Healthcare Informatics Logo

September 14th, 2013

By Mark Hagland

At Baystate Health, clinician and informatics leaders are moving forward in the elective early-term delivery area

The issue of unnecessary elective early-term deliveries of babies is receiving increasing attention in patient care organizations nationwide. This attention is not happening in a vacuum: an increasing number of reports and policy statements is backing up what has long been quietly understood among many clinicians: that many of the early-term deliveries are not only not medically indicated with regard to their timing, they can be detrimental to the health of babies and mothers.

For example, in June, the federal Centers for Disease Control and Prevention (CDC) reported that the total U.S. cesarean delivery rate reached a high of 32.9 percent of all births in 2009, having risen 60 percent from a rate of 20.7 percent of all births in 1996. Since 2009, though, that rate has remained unchanged.

Averting early-term births is a best practice that is actively being advocated by professional organizations involved in obstetrics, most notably the American College of Obstetrician Gynecologists (ACOG). Indeed, a joint March 2013 press release from ACOG and the Society for Maternal-Fetal Medicine (SMFM), noted that “The College and SMFM have long recommended that doctors not induce labor or perform cesareans before 39 weeks of pregnancy without a clear medical reason. A full-term pregnancy lasts 40 weeks. ‘Early-term’ deliveries are those that occur between 37 and 39 weeks of gestation.”

The March 21 press release noted that “Reducing the number of non-medically indicated early-term births and improving newborn outcomes is possible, according to The College and SMFM. Hospitals around the country have successfully lowered their rates of non-medically indicated early-term births by implementing policies to prevent them.”

One patient care organization in which active work to make such changes has been taking place is the four-hospital, Springfield-based Baystate Health.There, Andrew Healy, M.D., medical director of obstetrics, Daniel Grow, M.D., chair of the department of obstetrics and gynecology, and Peter St. Marie, the organization’s clinical research director, have been involved in groundbreaking work in this area. They have been leveraging PeriBirth EHR [electronic health record] solution from the Princeton, N.J.-based PeriGen, and have been using that data to optimize infant delivery at Baystate Health, where 45 physicians and 25 midwives perform 4,200 infant deliveries every year. Below are excerpts from the interview that Drs. Healy and Grow and Mr. St. Marie gave recently to HCI Editor-in-Chief Mark Hagland.

Can you share with me your overall strategy in pursuing this initiative?

Andrew Healy, M.D.: The background is that we certainly observed that babies were being delivered without a valid medical indication prior to 39 weeks; we also saw babies delivered after 39 weeks were being induced electively who weren’t favorable for induction, meaning that the mother’s cervix was closed. So often, the inductions of those patients would take more than one day—sometimes even three days—and many would fail, and they would have a cesarean section. And many of us have observed problems with the increase in c-sections across the country. And we’re also a teaching hospital, with both residents and medical students, so that element plays into the overall situation.

Daniel Grow, M.D.: And the problems with that are babies being born before they’re 39 weeks, and some of those babies go into the NICU [neonatal intensive care unit]. Furthermore, patients who are induced after 39 weeks’ gestation without a favorable cervix, spend a long time following induction, in labor, and often end up having a c-section. We found all those things unacceptable; so our goal was to eliminate uncalled-for inductions.

Healy: And in terms of process, we brought our faculty physicians and many community doctors into the process as well. And there have been statements from ACOG for more than 20 years, and evidence from recent publications around morbidity and mortality for these babies.

One of the things I’ve heard for years is that there is also a somewhat-hidden convenience factor for some of the physicians in this, having to do with disturbing their off-time with call or delivery.

Grow: You’re right about that, but some physicians will say, I’ve bonded with my patient and she wants me to be delivering while I’m on call or available. And sometimes there’s a reimbursement issue there, too, in terms of who gets paid for the cover delivery. So it’s complicated. But as the March of Dimes and ACOG and Leapfrog [the Leapfrog Group] have shown, babies are best delivered at term or when Mother Nature puts the mother into labor.

How did you reach consensus among all the physicians at Baystate for how to move forward with this?

Grow: Consensus is a strong word, right? So we had to show the doctors the data, and we showed that there were discrepancies in the induction rate between individual physicians and groups of physicians. So we shared the information with each group. There are four or five bigger groups, averaging about eight to 10 physicians each, here.

So you showed them the data in those groups, and got everyone to see what was going on?

Grow: Yes.

Healy: Yes, and there were marked differences between the groups.

So you had a discussion of the differences?

Healy: Yes, and because of Pete and because of PeriGen, we were able to review the inductions and determine whether the elective inductions were called for or not.

Peter St. Marie: So the baseline induction rate—any patient receiving pitocin—was about 16 percent. That figure represented all patients who received pitocin or induction, including for medically indicated reasons. The lowest group, two of them, had 16 percent receiving pitocin, and the highest rate was 40 percent. This is all patients receiving pitocin, either for medically justified reasons or for other reasons. So clearly, the 40-percent practices were performing non-medically indicated inductions. So Dr. Healy got the doctors together and said, let’s get together and define what “medically indicated” means, and it was us and the four leaders of the groups, plus anyone who wanted to join us. And we went diagnosis by diagnosis; there was some disagreement on different diagnoses, but Dr. Healy, through literature and skillful presentation skills, gained some consensus, and we narrowed it down to a list of reasons for medical indications. So we created a clinical pathway.

So to book an induction, the offices had to call a central scheduling office. And if they gave an approved medical indication for induction, they would book it. If they didn’t, they would have to call Dr. Healy directly. He’s a very nice guy; but he does have a backbone. So this was the basic plan. We’ve taken this a step farther since then, and now any patient who we discover is being induced without an approved medical indication, those charts are audited monthly, and if there’s a violation, that chart is sent to our peer review committee, where those charts are reviewed monthly, because some doctors have found a backdoor way to do that.

Healy: We went live in September 2011; and initially, I received a lot of phone calls. But as time has progressed and we’ve tracked the induction rate monthly, those phone calls have certainly decreased, and people have accepted the new policy. It’s not perfect, but I can tell you that many doctors are actually relieved about this, because the doctors can cite me or the hospital.

And so sometimes the convenience push is coming from the moms, too, of course?

Healy: Yes, and I’ll say that many of the pregnant moms simply don’t want to wait.

What is the rate now of early-term inductions that are not medically indicated?

Healy: The pre-39-week elective induction rate is now zero. And 40 percent was for all inductions in that group.

What percentage of the 40 percent figure of early-term inductions was not medically indicated, do you think?

Healy: I don’t have that number off the top of my head, but I suspect that it could have been half of that percentage, because I would guess that 20 percent were medically indicated. And I can add that when we looked at the numbers for the number we presented at ACOG, our elective induction rate was 6 percent and fell to 2 percent after we instituted the policy.

How long did it take to fall from the 6 percent to the 2 percent?

Healy: From January 1, 2009 to August 1, 2011.

So over two-and-half years?

Well, the post-policy period was actually form September 2011 to June 2012.

So that rate fell relatively quickly, then?

Healy: Yes, because these hard stops were put in place; if they didn’t satisfy certain criteria, they weren’t allowed to book the induction.

What is the core of what PeriGen does?

Healy: What PeriBirth does is it’s the EHR, which and allows us to track each delivery and induction.

St. Marie: And attempting a study based on the old paper records is incredibly time-intensive. Here, we can stratify all patient types and situations.

Grow: We can query the record for anything we want. The big advantage of having PeriBirth and any EHR, is that you can pull all the records quickly and in a systematic way. I pulled all the delivery reports for all the deliveries for the whole time period, and was able to pull that into some software and then analyze the data any way we wanted, and for even making the case at the beginning to even make the case for doing this.

What would you say to CIOs and CMIOs about how to help facilitate this kind of initiative?

Grow: From the clinical side, there are cases where they’re forced to enter data.

Healy: They’re forced to enter data into certain fields.

St. Marie: That’s a big part of getting complete data into the medical record. And the other big part of PeriBirth is, I can do a query of the database to look at all deliveries between any two dates, and can get all the information, basically.

Healy: And the physicians love that we can track things: how often are docs doing an episiotomy, or how often is pitocin being given? So we can track all things that are important to us, and can educate the doctors on things.

St. Marie: And it’s relatively easy for me to give Dr. Healy a timeframe, and to pull any kind of indicator, and can merge everything together and do an analysis on it.

Grow: So from a CQO type of perspective, we can create quality dashboards that are very specific and that save money and improve care, so that’s very valuable. So from a departmental perspective, we can run report cards, if you will, to drill down from who’s delivering excellent quality and who isn’t.

St. Marie: From the IT perspective, once we got the software, the codes able to run to do the analysis, we were able to reuse that every month, and we’re able to print out a monthly report that Dr. Healy presents at grand rounds, and that is re-created in comparison with previous months. And in a few hours, we’ve analyzed all the deliveries for the month, and have specifics on inductions, elective inductions, etc.; and the only way we’re able to do that is because we can query PeriBirth; and I use software that does ODBC queries, to talk to the server.

Grow: So what’s very useful are the quarterly visits from PeriGen that have a standard report of about 100 quality metrics that we can track internally, and so one of them is frequency of blood transfusion; another is frequency of third- and fourth-degree episiotomy; another is frequency of perfect documentation from our nursing team. So we get that quarterly and get benchmarking across every organization that uses PeriGen. They have approximately 400,000 deliveries in their database; I think there are about 30 other organizations using it.

This is part of the broader shift towards evidence-based, consensus-driven, medical practice, isn’t it?

Healy: I would agree that there’s definitely a movement towards evidence-based care, and that’s been going on for a while now. There are two things I would stress: first, you need to be able to establish the evidence, and systems like PeriBirth do that. And second, you need to be able to measure deviations against standards of care; and again, PeriBirth does that, and you need to be able to address who is deviating, and address that.

St. Marie: And PeriBirth is our inpatient OB EMR; and the hospital has a separate EMR for everyone else. That is not as easy to query, as the data in that database is not updated as frequently as the PeriBirth data. With PeriBirth, I can get data for patients still in the hospital and orders just placed. With the inpatient EMR, it’s nearly historical data, and requires a much more difficult, proprietary method. So that’s the great thing about PeriBirth for me; I can write a few lines of code to address a query, and have the information back to them within a half-hour.

Healy: And once an acceptable standard has been established, we can use PeriBirth for that purpose. What is the correct dosage for a particular medication in a particular situation, for example? Standardizing care is very important.

Grow: And in this study, putting this policy in place, hard stops for induction of labor, reduced our NICU admission rate significantly, for patients over 37 weeks gestation, from 3 percent to 2 percent, so a reduction of a third. So the NICU admission rate dropped by over a third in that group.

Source URL: http://www.healthcare-informatics.com/node/18500?page=0

23 Aug

On the Leading Edge of Labor and Delivery Operations: Norwalk Hospital Implements Real-Time Delivery Room Surveillance

Healthcare Informatics Logo



February 15, 2013

By Rajiv Leventhal

While the ability to see new lives enter the world every day is a blessing, being an obstetrician (OB) or a delivery room nurse can be quite stressful. Many are working in understaffed environments, are tired, new, or just having a bad day, and having access to an “expert opinion” right at the bedside to help them make crucial decisions can be very beneficial.

Enter PeriGen, a Princeton, N.J.-based provider of fetal surveillance systems employing patented, pattern-recognition and obstetrics technologies that empower perinatal clinicians to make confident, real-time decisions about the mothers and babies in their care. In January 2012, the 328-bed Norwalk Hospital in Connecticut decided to implement PeriCALM Tracings, which provide real-time, automated assessments of labor progress and fetal heart rate patterns.

After a study—released in January 2013 and validated by the National Institutes of Health (NIH)—revealed that the PeriCALM software can be used to efficiently evaluate stored fetal heart rate data retrospectively as well as to screen in real time when an expert is not continuously evaluating the fetal heart rate, Healthcare Informatics Assistant Editor Rajiv Leventhal had a chance to speak with Yoni Barnhard, M.D., Chairman of the Department of Obstetrics and Gynecology at Norwalk Hospital, about the advantages of using software in delivery rooms to screen in real time.

How has the implementation of PeriCALM affected OBs and delivery room nurses?
They love the system, across the board. First, it’s absolutely improved communication between nurses and physicians. It’s given the nurses a more credible voice in discussions with physicians about the analysis of a fetal heart rate tracing. Across the country, especially in community hospitals like ours, physicians come to work in the morning, start the induction process on their patient, go to the office, and become reliant on whoever is in the hospital (nurses, residents, hospitalists, etc.) to help them analyze what is going on.

Now, they can see in real time what is happening, even though they are in the office. And the nurses are looking at the same tracing. If there is any discrepancy with the interpretation of the tracing, PeriCALM is used to provide a real-time assessment of what is happening. This means there are far fewer instances where nurses have felt the need to go up the chain of command in order to just be able to get a physician’s attention about a concern they have. This has absolutely changed the safety culture in terms of labor and delivery. We had done a safety culture survey before we implemented PeriCALM and in the next few months, we’ll do another one. I think we will see that it has changed the nurse’s perception of perinatal safety.

And the alert system allows for enhanced situational awareness on the labor floor. In the central command area, we have a large flat screen on the wall and you can see all of the fetal heart tracings on there. Everyone can see the same thing, and if there is anything in any room out of the ordinary in any tracing, you get an alert. And anyone can respond to it who is sitting there. I think the nurses feel they are not the only ones looking at a patient’s tracing. If they are doing something else at that moment, there is a whole team of people looking, making the data much more objective. Nurses are very happy with that.

How do the PeriCALM alerts specifically work?

PeriCALM is built on the National Institute of Child Health & Human Development (NICHD) criteria for fetal heart rate tracings, which looks at five main components: baseline fetal heart rate; presence or absence of accelerations (increases in fetal heart rate, which are reassuring); presence of decelerations (which may be of concern); variability (how fetal heart rate changes from one beat to the next); and uterine contractions. If any of these five parameters are abnormal, as defined by national guidelines, nurses will get an alert. Normally, we would be able to see most of these things, but wouldn’t be able to respond immediately. That is the big change—there is no delay in recognition that something needs to be addressed. It’s enabled us in a more timely fashion to recognize that there is a potential problem.

What are some other advantages of implementing this kind of delivery room software?

In 25 years, I have never been in a hospital that had an alert system that was both auditory and visual. We’ve always had the most traditional type of fetal heart monitoring systems, which is what most hospitals have—just a record of the fetal heart rate. But if it is at night, and the hospital is not ideally staffed, nurses might see something that is not quite right but cannot respond immediately due to being so busy. Now, with PeriCALM, you have to respond because the system will keep alerting you. Every alert requires a response, and all of that information goes into the electronic medical record (EMR). A physician, for example, might get a pop up on his or her screen saying that pitocin should not be administered due to a patient’s category II tracing. So if the physician attempted to increase the pitocin, he or she would be unable to do so. It will make what we do much more evidence-based and it will help doctors and nurses follow best practice guidelines at the most critical decision points. The main reasons for medical errors revolve around communication errors and bad physician decision making. These tools will help prevent errors and are part of an overall comprehensive safety plan.

What advice would you give to other hospitals concerning fetal surveillance systems?
When you take state-of-the-art technology and put it on a labor floor, it makes a significant difference. It improves documentation, patient care, and enhances safety. If I can do one thing in this country to change care with labor and delivery units, it would be to put PeriGen units in. And I have no financial incentive to say that.
Source URL: http://www.healthcare-informatics.com/article/leading-edge-labor-and-delivery-operations-norwalk-hospital-implements-real-time-delivery-ro

Download Article

21 Feb

Software Assistants for Doctors Are Making Progress

NY Times Bits story published 2/17/2013 by Steve Lohr

Doctors have long been in the high-stakes information management business. They must quickly sort through a patient’s symptoms, comments, test results, records and history to come up with a diagnosis. The physician brings to each diagnostic encounter a storehouse of knowledge and experience, all that he or she has read and learned over years.

The information overload for doctors is only growing worse. Medical information is estimated to be doubling every five years, and surveys show most doctors can find only a few hours a month to read medical journals. So it is not surprising that automated assistance for doctors has been pursued by researchers and companies for many years. Decision-making in medicine, after all, involves not just time and money, but also human lives.

A worthy goal, but a frustratingly difficult one. Yet in the last few years, real progress is being made in what is called “clinical decision support” technology. And the story in medicine is the same as in so many applications of modern computing: advances in sensors for measuring, calculating power and artificial-intelligence software are opening the door to a new generation of smarter tools.

The marquee example is I.B.M.’s Watson. After defeating human “Jeopardy” champions, the clever question-answering computer has moved on to medicine. It is working with oncologists at Cedars Sinai Cancer Institute in Los Angeles, and Watson is being trained as a medical student at the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University.

But there are a flurry of other, smaller-scale examples from teams in universities and start-ups. The promising work of one such company, PeriGen, was published recently in the American Journal of Obstetrics and Gynecology and is being presented on Monday in San Francisco at the annual clinical meeting of the Society of Maternal Fetal Medicine.

The company, based in Princeton, N.J., specializes in fetal monitoring technology. The research being presented is an assessment by the federal National Institutes of Health, and the company’s software grew out of years of work by physicians and scientists at McGill University in Montreal. The team includes computer scientists, engineers, mathematicians and statisticians, and has been led by Emily F. Hamilton, an obstetrician and gynecologist who is senior vice president for clinical research at PeriGen, and an adjunct professor at McGill.

The software assessed in the N.I.H. study, called Pericalm Patterns, collects and analyzes the data from fetal heart monitors. Traditionally, doctors see the output of fetal heart monitors as wavy lines — called tracings — printed on paper that scrolls off the machine (roughly similar to the scribblings of lie-detector tests). Physicians are looking for patterns that might suggest the baby is in distress, perhaps suffering from oxygen deprivation. If so, they typically order a Caesarean section, to get the baby out quickly.

The trouble for obstetricians, Dr. Hamilton said, is that most babies, as they are about to be born, exhibit some unusual heart rate patterns. “You’re looking at all those tracings, and you’re applying rules of thumb and a lot of judgment,” she said in an interview over the weekend. “The challenge is to distinguish what is critical from what is just distracting.”

The N.I.H. assessment concluded that the analysis of three human experts agreed with that of the company’s software 97 percent of the time. The data used for this phase of the study was 100 tracings in the final hour before a baby was born. The technology can be used for research to identify characteristic tracing patterns, and in real time to give physicians alerts.

A second phase of the N.I.H. study, which is already under way, will involve 5,000 tracings — or about 20 miles of paper. “You can see why that cannot be done by humans. But we’ve reached the point now where we can use modern techniques of statistics and computing to really advance the science.”

Better science, Dr. Hamilton said, should pave the way for better, more consistent decision-making in obstetrics. On the one hand, she said, that should mean faster, clearer signals of a baby in distress. On the other hand, Dr. Hamilton said, it should also reduce unnecessary Caesarean-section deliveries. Faced with uncertainty, physicians tend to order a Caesarean to remove any risk to the baby, as well as the risk of malpractice suits.

While leveling off, the rate of Caesareans — at about 32 percent of all births in the United States — is twice as high as it was two decades ago.

Full Article

Download PDF

13 Jun

High uterine contraction rates in births with normal and abnormal umbilical artery gases

Emily F. Hamilton, Philip Warrick, G. Eric Knox, Daniel O’Keeffe, Thomas J. Garite


Objective: To determine if the incidence of high contraction (HC) rates and associated decelerations were different in term births with metabolic acidemia (MA) compared to those with normal gases (N) over the last 4 h of labor.

Methods: MA included 316 babies with cord base deficits (BD) over 12 mmol/L N – 3,320 babies with BD under 8 mmol/L. HC rates were defined as >5/10 min.

Results: One or more episodes of HC occurred in 43.7% of MA and 36.6% of N. (p = 0.015) In both groups the HC rates rose from about 1 in 30 patients at the beginning to 1 in 7 to 9 patients at the end. MA showed a different transition of the deceleration response over time. At the beginning the average ratio of decelerations to uterine contractions was similar in both groups but over the final 140 min MA showed a consistently higher ratio.

Conclusions: Although HC rates were more frequent in the MA, it was not uncommon in N. On average MA showed more decelerations at every level of contractions and had a persistently higher level of decelerations per contraction for more than 2 h before birth.


Cite this article as: Hamilton E, Warrick P, Knox E, Keeffe,DO, and Garite T. High uterine contraction rates in births with normal and abnormal umbilical artery gases. J Maternal Fetal Neonatal Med. 2012 Jun 13.

04 Jun

Outcomes associated with a structured prenatal counseling program for shoulder dystocia with brachial plexus injury

Mary Veronica Daly, MD, Christina Bender, MSN, Kathryn E. Townsend, JD, Emily F. Hamilton, MD


Objective: We examined outcomes associated with a novel program to identify patients at high risk for shoulder dystocia (SD) with BPI.

Study Design: The program included a checklist of key risk factors and a multifactorial algorithm to estimate risk of SD with BPI. We examined rates of cesarean and SD in 8,767 deliveries by clinicians enrolled in the program and in 11,958 patients of clinicians without access to program.

Results: Key risk factors were identified in 1,071/8767(12.2%) of mothers of whom 40/8767(0.46%) obtained results in the high risk category. The rate of primary cesarean rate was stable (21.2% to 20.8% P= 0.57). SD rates fell by 56.8%. (1.74% to 0.75% P=0.002). The rates of SD and cesarean birth showed no changes in the group without access to the program.

Conclusions: With the introduction of this program, overall SD rates fell by more than half with no increase in the primary cesarean rate.

Cite this article as: Daly M, Bender C, Townsend K, and Hamilton E. Outcomes associated with a structured prenatal counseling program for shoulder dystocia with brachial plexus injury. AJOG 2012 June 04.

Full Article

26 Jul

Shoulder Dystocia: Neither Predictable nor Preventable? Not Anymore!

In the Summer, 2011 issue of MD Advisor, MD Advantage shares the results of a three-year trial of the PeriCALM® Shoulder Screen™ tool, a web-based application to identify mothers at greatest risk of shoulder dystocia. Using PeriCALM Shoulder Screen, MD Advantage’s insured physicians were able to reduce the incidence of shoulder dystocia by 50%, with no increase in the primary cesarean rate– disproving the myth that shoulder dystocia is an unpredictable and unpreventable occurrence. The article, “Shoulder Dystocia: Neither Predictable nor Preventable? Not Anymore!” is based on an interview with Emily Hamilton, MD, SVP of Clinical Research at PeriGen, Mary Veronica Daly, MD, Practicing Obstetrician at Lifeline Medical Group and Atlantic Health, and Robert B. Goley, SVP of Claims and Risk Management at MD Advantage.

Click here to read the rest of this article.