declining fertility

Report
Physician Financial Incentives and
Cesarean Section Delivery
Gruber and Owings
Problem of Testing PID
Problem of Testing PID

Using fee changes:


Identification problem: it can’t identify supply response
When fee goes up
 demand less
 supply more (subs. effect) / supply less (income effect)
3
Problem of Testing PID

Using ob/gyns density:
Omitted variable problem: COV(X,U) ≠0
 Y
= XB
+ U
(c-section) (density) (local coinsurance rate)

生產方式
自然產
自然產是指在不使用器械或開刀的情況下,產婦
經由陰道生產分娩胎兒的方式,也就是將懷孕的
產物(包括胎兒、羊水、胎盤及胎膜)剝離子宮
而經陰道娩出的過程。
 剖腹產
剖腹產是婦科常見的分娩手術,以手術方式切開
腹壁和子宮壁,將體重達到500克以上的胎兒娩
出,並且於胎兒取出後,直接清理子宮內胎盤及
胎膜等,術後縫合傷口。

選用剖腹產原因
骨盆狹小、胎兒過大
 胎兒窘迫
 前置胎盤
 胎位不正(尤其為初產婦 )
 胎盤早期剝離
 臍帶脫出


引產失敗(即生產過程中難產)
自然產、剖腹產優缺點
自然產
自然產的子宮感染機率、傷口感染機率、嚴重出
血機率、泌尿道感染機率皆較低;產後恢復快,
產後改善痛經效果好。
 剖腹產
剖腹產將來引起尿失禁、骨盆腔損傷的機率低。

各國剖腹產率
台灣:過去十年來介於32-34%間
 世界各國








美國:24.4%(2001);30﹪(2006)
英國:22.6%(2001);23.3﹪(2006)
法國:17.8 %(2001)
日本:21.4% (2005)
韓國:39.5% (2001);35.2﹪(2006)
最高的幾個國家:阿根廷約六成,中國大陸約五成
WHO 1985年建議剖腹產率介於10~15%,但
目前世界各國的剖腹產率大多高於這個水準
Cesarean delivery in the US
There is no consensus view as to the cause of this rapid
increase in cesarean utilization.
 The most frequently cited explanation is the
introduction of technologies for diagnosing fetal
distress.
 Another important cause of c-section adoption was
thought to be the legal environment.

表1. 歷年生產案件和健保支付點數
年度
支付點數
自然產
剖腹產
a,b
申請案件佔率
自選剖腹
1998
案件數
自然產
剖腹產
自選剖腹
64.68
32.76
1.70
253,600
1999
|
|
|
64.89
32.48
1.68
267,218
2000
17,420
32,330
17,420
64.35
32.48
1.99
286,714
2001
17,910
33,280
17,910
64.59
32.09
2.37
245,271
2002
|
|
|
64.60
32.03
2.47
237,055
65.36
31.45
2.23
222,364
2003
2004
18,268
33,969
18,268
66.23
30.97
1.73
214,541
2005
33,969
33,969
33,969
65.55
31.69
1.67
204,406
2006
36,086
36,086
15,188
65.13
31.90
1.93
202,764
2007
|
|
|
64.26
32.64
2.15
200,384
資料來源:根據中央健康保險局支付標準歷年公告加以彙整。
a. 不同院所層級的生產給付點數有所差異,本表僅列出醫學中心支付點數。
b. 自2003年4月起,新增剖腹產後自然產(VBAC)論病例計酬項目,比照剖腹產支付點數。
圖1:歷年生產案件剖腹產率變化
36.00%
35.00%
34.00%
33.00%
32.00%
31.00%
30.00%
29.00%
28.00%
27.00%
26.00%
1998
1999
2000
2001
2002
剖腹產(全部)
2003
剖腹產(首胎)
2004
2005
2006
2007
Financial incentives for cesarean delivery

But, why the cesarean delivery reimburses more then
normal childbirth?
The reason
1. Cesarean delivery is more difficult then vaginal
delivery.
2. Cesarean delivery is much more efficient.
3. Cesarean delivery is also more costly than vaginal
childbirth.
Research Question

They exploit a plausibly exogenous change in the
financial environment facing obstetrician/gynecologists
during the 1970s: declining fertility in the United
States.

The interest question is: In states where fertility was
falling the most, did cesarean delivery rise the
most?
The Model (McGuire and Pauly)
The data

The data source is the National Hospital Discharge
Survey (NHDS) conducted annually since 1965 by the
National Center of Health Statistics (NCHS).

This is a nationally representative survey of over 400
hospitals, and it contains information on approximately
200,000 discharges in each year.
The data

The survey collects information on some limited
demographic characteristics of the patient, features of
the hospital, primary and secondary diagnoses, and
primary and secondary procedure use.

The authors use information on patients with a primary
diagnosis of childbirth in the years 1970-1982.
Estimation Model
Basic results

A 10% increase in the ob/gyn density raises the
probability of a c-section by 0.6%

A 10% fall in the fertility rate raises the probability of a
c-section by 0.97%
25

How large are these effects?
10% decline in fertility rate 5% drop in income
1% rise in c-section usage 1.68 more c-section per
ob/gyns per year ($943 ≒0.5% increase in income)

The c-section offsets only a very small part of the
negative income shock facing ob/gyns
26
Specification checks

The CS is correlation with the riskiness of birth

The CS is correlation with the birth order.

The CS is could be a function of changing insurance
coverage of mothers.
Specification checks
Specification Checks
After including birth severity controls, the coefficient
on fertility is unchanged
 Including birth order information has little effect on the
results
 Even excluding 5 outlying states, the correlation
between fertility and c-section remains robust
 The inducement effect is higher for private insured
women than uninsured (1.15% vs 0.65%)

31
Asymmetric Response?

similar documents