Evaluating Healthcare Policies to Reduce Rates of Cesarean Delivery


Cesarean delivery is a method of childbirth in which a surgeon cuts through a pregnant woman’s abdomen and uterus to deliver the baby. The more natural method of childbirth is vaginal delivery, in which the baby leaves the mother’s uterus through her vaginal canal.

Ideally, cesarean delivery would only be used when vaginal delivery would endanger the life or health of the child or mother. It involves major abdominal surgery, which is accompanied by much greater risks for both mother and child than vaginal delivery. Cesarean delivery also costs about 50 percent more.

Over the last 40 years, the United States rate of cesarean delivery has increased dramatically, from 4.5% in 1965 to 32.8% in 2012. Many of these cesarean deliveries, some researchers say 50 percent of them, are unnecessary. In 2009, childbirth related hospitalizations accounted for 7.6 percent of all inpatient costs, totaling $27.6 billion.

The challenge of reducing the cesarean delivery rate has been recognized by numerous researchers for years. For the first time, in research conducted for the Washington State, Alan Mills, FSA MAAA ND, a research actuary, and his colleagues reproduced this part of the United States healthcare system in a simulation model to allow the stakeholders, including health agencies, insurers, clinicians, and legislators, to test their assumptions on the model to find the right solutions.


The researchers approached the problem on an individual state level as states differ in terms of birthing environment, including healthcare infrastructure and Medicaid conditions. At this stage of the project, they looked at four states: Washington State, Illinois, New York, and West Virginia, because they wanted a broad range of cesarean delivery rates.

They decided to choose the agent-based modeling method because it allowed the modelers to reflect the behavior of many heterogeneous agents acting independently. Moreover, agent-based modeling allowed them to reflect the network effects and feedback, when women and obstetricians make their decisions based on what their friends and colleagues do and previous experiences.

Regional Healthcare Simulation Model Animation

Model Animation on the Region's Map

Agent Behaviour in Healthcare Simulation Model 

General Agent Behaviour in the Model

The model included the following types of agents:

  • Women 
  • Obstetricians 
  • Certified nurse midwives (CNMs) 
  • Licensed midwives (LMs) 
  • Hospitals 
  • Health insurers 
  • Medicaid

The researchers parametrized the agents with real world data imported from Excel. For the private information, they generated synthetic data based on the real data.

It was important to simulate the key behaviors of the agents correctly to reflect the real decision making process. The simulated behaviors included women choosing birth attendants (obstetricians, CNMs, or LMs), birth attendants choosing hospitals of association, delivery attendance (vaginal or cesarean), the process of submitting a claim to insurers/Medicaid, and payment.

All of the behaviors were based on the same template that had ten main components (see picture). For instance, women chose their birth attendants based on their childbirth month (input data), motivation (desired birth attendant type), and attributes (geographical location).

The model had an interactive interface so that the users could play with parameters (motivations, geographical distributions, etc.), run experiments, and see how different policies would affect system behavior. The results window included financial and population outcomes so the users could easily see and compare the effectiveness of various strategies. The model also included visualization on the map, where all of the agents were animated, and the problem was given a geographical perspective.


With AnyLogic, the researchers created a game oriented, simulation based, learning environment for the stakeholders to better investigate the problem and test possible strategies. The training sessions that used the simulation model significantly increased the stakeholders’ understanding of the problem and could lead to the development of better policies to decrease the unnecessary cesarean delivery rate.

The primary results of the experiments showed that the previously suggested healthcare provider payment reform would not work alone, and the stakeholders would need a multifaceted strategy. Also, if retooling the workforce was chosen as part of the strategy, it would take a lot of time.

This was the first simulation model in the world addressing the high cesarean delivery rate issue on a regional level.

Watch the video of Alan Mills, FSA MAAA ND, presenting this case study:

More Case Studies

  • 医薬品流通倉庫のモデル化
    数十億ドル規模売上を計上する、医薬品流通およびロジスティクス企業大手のカーディナル・ヘルス社(Cardinal Health)は、商標医薬品、ジェネリック医薬品、一般用医薬品(薬局で処方箋が無くても購入できる薬)、健康と美容アイテムおよび自社ブランド品を管理しています。彼らは、医薬品を取り巻く複雑な倉庫の流通問題に取り組みます。ブライアン・ヒース(カーディナル・ヘルス社上級解析ディレクター)とAnyLogicソフトウェアの経験ある担当者は、様々なビジネス問題を解決するためにエージェントベースのモデリングを使用し、年間300万ドル以上の経費を節減しました。
  • ハイブリッド・シミュレーションによる医療サポート-脳卒中専門救急車(Mobile Stroke Units)
    脳卒中で高度障害が起きたときの治療とリハビリの高コストの負担は老齢人口の増加でますます増えています。血栓症の多くは脳卒中を発症し、発症から4.5時間以内に血栓溶解の治療が必要になりますが、現在の搬送や病院管理では対応しきれていないのが現状です。そこで、脳卒中専門救急車(Mobile Stroke Units)が改善案として提案されました。
  • 医療ルーチンデータのシミュレーション・モデリング
    医療の専門家による様々な意思決定には、プランニング、テストおよびアセスメントツールを必要とします。医療の複雑な構造、相互作用およびプロセスは、常に変化と革新を繰り返し、課題が絶えることはありません。社会保険オーストリア協会(AASI)と提携するDWHシミュレーションサービスおよびウィーン工科大学のPatrick Einzinger氏およびChristoph Urach氏は、クリティカルな将来の意思決定の目的で、医療データを解析する機会を得ました。
  • 産科病棟シミュレーション
    このモデルは、建設中の病院の産科病棟をシミュレートします。モデルの目的は、新設病棟におけるリソース、キャパシティーおよび業務をシミュレートすることです。プロジェクトはスウェーデン ストックホルムのカロリンスカ大学病院で実施されました。
  • 入院患者治療キャパシティーの評価
  • 透析患者トータルケアのニーズ
    スウェーデン 首都ストックホルムは、諸国、諸地域と同様、様々な患者グループへのヘルスケアの必要性があります。各グループは独特な特徴、特性および課題を合わせてサブ集団と見なします。今回ご紹介するシミュレーション・プロジェクトは、透析患者(医療施設の来院頻度が高いグループ)に着目しました。
  • Disaster Response Applications Using Agent-Based Modeling
    In an effort to find practical operational solutions for response to an unexpected crisis or natural disaster, Battelle, world’s largest, non-profit, independent R&D organization, needed to test the effectiveness of a 48 hour shelter-in-place order for an Improvised Nuclear Device scenario. The goal was to reduce radiation dosages received during an uncoordinated mass evacuation, by comparing immediate evacuation and shelter-in-place order.
  • Shaping Healthcare Policy Using Simulation
    An initiative by the Department of Mechanical and Industrial Engineering at the University of Toronto, the Centre for Research in Healthcare Engineering (CRHE), was in response to the immediate and compelling desire for efficiency and quality improvements in the Canadian healthcare system.
  • An Agent-Based Explanation for SPMI Living Situation Changes
    Over the past 60 years, the number of Severely and Persistently Mentally Ill (SPMI) patients in the US living in the community increased. Yet a growing minority of people with severe illness are worse off because they are homeless or incarcerated. In this case study, IBM Global Research and Otsuka Pharmaceuticals used an agent-based approach to model these remarkable swings.
  • Outpatient Appointment Scheduling Using Discrete Event Simulation Modeling
    Indiana University Health Arnett Hospital, consisting of a full-service acute care hospital and a multispecialty clinic, faced poor statistics because the number of no-show patients (those who don’t show up for their scheduled appointments) rose dramatically to 30%. This was primarily connected to the fact that clinic schedules were driven by individual preferences of the medical staff, which led to increased variations in scheduling rules. To eliminate the problem, the client wanted to develop a scheduling methodology that would benefit the clinic, doctors, and patients.