Assessing	the	Brazilian	Surgical	System	with	Six	Surgical	Indicators:	A	Descriptive	and	 Modeling	Study	
Running	Head:	Surgery	in	Brazil	 Authors:	
Benjamin	B.	Massenburg	BA*1,2,3	 Saurabh	Saluja	MD	MPP*2,3,4	 Hillary	E.	Jenny	MPH1,2,3	 Nakul	P.	Raykar	MD	MPH2,3,5	 Josh	Ng-Kamstra	MD	MPH2,3,6	 Aline	G.	A.	Guilloux	MsC7	 Mário	C.	Scheffer	PhD7	
John	G.	Meara	MD	DMD	MBA2,3	 Nivaldo	Alonso	MD	PhD†8	
Mark	G.	Shrime	MD	MPH	PhD†2,9	 *	Co-first	authors	 †	Co-last	authors	 Affiliations:	 1Department	of	Medical	Education,	Icahn	School	of	Medicine	at	Mount	Sinai,	New	York,	USA	 2Program	in	Global	Surgery	and	Social	Change,	Harvard	Medical	School,	Boston,	USA	 3Department	of	Plastic	and	Oral	Surgery,	Boston	Children's	Hospital,	Boston,	USA	 4Department	of	Surgery,	Weill	Cornell	Medicine,	New	York,	USA	 5Department	of	Surgery,	Beth	Israel	Deaconess	Medical	Center,	Boston,	MA,	USA	 6Department	of	Surgery,	University	of	Toronto,	Toronto,	Canada	 7Departamento	de	Medicina	Preventiva,	Faculdade	de	Medicina	da	Universidade	de	São	Paulo,	São	 Paulo,	Brasil	 8Departamento	de	Cirurgia	Plástica,	Faculdade	de	Medicina	da	Universidade	de	São	Paulo,	São	Paulo,	 Brasil	 9Department	of	Otolaryngology,	Massachusetts	Eye	and	Ear,	Boston,	USA	 	
1	 	

Financial	Disclosures:	There	was	no	funding	for	this	study.	No	authors	have	a	financial	interest	in	any	of	 the	products,	devices,	or	drugs	mentioned	in	this	manuscript.	Dr.	Shrime	reports	speaking	fees	from	 Ethicon	in	January	2014,	outside	the	submitted	work.	 Presentations:	This	paper	will	be	presented	at	the	2016	Annual	CUGH	Global	Health	Conference	in	San	 Francisco,	CA	on	April	9-11,	2016.	 Author	Contributions:	 BBM:	literature	search,	figures,	study	design,	data	collection,	data	analysis,	data	interpretation,	writing	 SS:	literature	search,	figures,	study	design,	data	collection,	data	analysis,	data	interpretation,	writing	 HEJ:	literature	search,	data	analysis,	data	interpretation,	writing	 NPR:	literature	search,	data	analysis,	data	interpretation	 JNK:	literature	search,	data	analysis,	data	interpretation,	writing	 AGAG:	literature	search,	data	analysis,	data	interpretation	 MCS:	literature	search,	data	analysis,	data	interpretation	 JGM:	study	design,	data	interpretation,	writing	 NA:	study	design,	data	analysis,	data	interpretation,	writing	 MGS:	study	design,	data	analysis,	data	interpretation,	writing	 	 Corresponding	Author:		 Mark	Shrime	MD	MPH	PhD		 641	Huntington	Avenue	 Boston,	MA	02115	 +1	617	432	6942	 shrime@mail.harvard.edu		
2	 	

	 ABSTRACT:	 Background:	Brazil	boasts	a	health	scheme	that	aspires	to	provide	universal	coverage,	but	its	surgical	 system	has	rarely	been	analyzed.	In	an	effort	to	strengthen	surgical	systems	worldwide,	the	Lancet	 Commission	on	Global	Surgery	proposed	collection	of	six	standardized	indicators:	two-hour	access	to	 surgery,	surgical	workforce	density,	surgical	volume,	perioperative	mortality	rate	(POMR),	and	 protection	against	impoverishing	and	catastrophic	expenditure.	This	study	aims	to	characterize	the	 Brazilian	surgical	health	system	with	these	newly	devised	indicators	while	gaining	understanding	on	the	 complexity	of	the	indicators	themselves.	 Methods:	Using	Brazil’s	national	healthcare	database,	commonly	reported	healthcare	variables	were	 used	to	calculate	or	simulate	the	six	surgical	indicators.	Access	to	surgery	was	calculated	using	hospital	 locations,	surgical	workforce	density	was	calculated	using	locations	of	surgeons,	anesthesiologists	and	 obstetricians	(SAO),	and	surgical	volume	and	POMR	were	identified	with	surgical	procedure	codes.	The	 rates	of	protection	against	impoverishing	and	catastrophic	expenditure	were	modelled	using	cost	of	 inpatient	hospitalization	and	a	gamma	distribution	of	incomes	based	on	GINI	and	GDP/capita.		 Findings:	In	2014,	SAO	density	is	34·7/100,000	population,	surgical	volume	is	4,433	procedures/100,000	 people	and	POMR	is	1·71%.	79·4%	of	surgical	patients	are	protected	against	impoverishing	expenditure	 and	84·6%	were	protected	against	catastrophic	expenditure	due	to	surgery	each	year.	Two-hour	access	 to	surgery	was	not	able	to	be	calculated	from	national	health	data,	but	a	proxy	measure	suggested	that	 97·2%	of	the	population	has	two-hour	access	to	a	hospital	that	may	be	able	to	provide	surgery.	 Geographic	disparities	were	seen	in	all	indicators.	 Interpretation:	Brazil‘s	public	surgical	system	meets	several	key	benchmarks.	Geographic	disparities,	 however,	are	substantial	and	raise	concerns	of	equity.	Policies	should	focus	on	stimulating	appropriate	 geographic	allocation	of	the	surgical	workforce.	In	some	cases,	where	benchmarks	for	each	indicator	are	 met,	supplemental	analysis	can	further	inform	our	understanding	of	health	systems.	This	measured	and	 systematic	evaluation	of	surgical	systems	should	be	encouraged	for	all	nations	seeking	to	better	 understand	their	surgical	systems.		 Funding:	There	was	no	funding	for	this	study.	
3	 	

	 What	this	paper	adds:	 What	is	already	known	on	this	subject:	
• There	is	a	substantial	lack	of	surgical	monitoring	and	evaluation	worldwide.	 • The	Lancet	Commission	on	Global	Surgery	(LCoGS)	proposed	the	six	surgical	indicators	used	in	
the	present	study.		 • Geographic	disparities	in	the	trainee	and	physician	workforce	have	been	reported,	though	no	
other	studies	were	identified	that	characterized	other	elements	of	the	surgical	system	of	Brazil.	 What	this	study	adds:	
• This	is	the	first	comprehensive	and	critical	analysis	of	the	surgical	indicators	proposed	by	LCoGS	 in	any	country,	demonstrating	that	the	collection	of	surgical	system	indicators	is	possible	where	 national	health	data	is	recorded.	
• This	study	provides	an	in-depth,	multi-factorial	assessment	of	Brazil’s	surgical	system	while	also	 identifying	and	quantifying	geographic	disparities	that	exist.	
• Recommendations	for	the	LCoGS	surgical	indicators	and	for	the	Brazilian	surgical	system	are	 summarized	in	Table	3.		
4	 	

	 INTRODUCTION:	
In	a	2015	report,	the	Lancet	Commission	on	Global	Surgery	(LCoGS)	estimated	that	over	5	billion	people	 lack	access	to	safe,	affordable,	and	timely	surgical	care	worldwide	and	that	an	additional	143	million	 surgical	procedures	are	needed	annually	to	meet	global	needs.[1	2]		Recognizing	the	critical	role	of	 measurement	in	understanding	and	tracking	the	state	of	a	surgical	system,	LCoGS	proposed	six	 standardized	indicators,	with	associated	benchmarks	for	two	of	them,	(Table	1)	and	advocated	for	the	 incorporation	of	these	indicators	into	existing	mechanisms	for	health	system	assessments.	While	these	 indicators	were	developed	through	an	expansive	and	iterative	expert	consultation,	they	have	not	yet	 been	comprehensively	applied	at	the	national	level.	
Brazil,	the	world’s	fifth	most	populous	country,	aspires	to	provide	universal	health	coverage.	As	such,	it	 is	an	optimal	pilot	site	for	studying	these	indicators.	The	country	is	comprised	of	five	regions,	which	are	 further	divided	into	26	states	and	a	federal	district.	In	the	1980s,	Brazil	was	the	world’s	most	unequal	 country[3]	with	the	South	and	Southeast	regions	representing	a	significantly	wealthier	population	than	 the	rest	of	the	country.	Nevertheless,	the	nation	has	experienced	rapid	economic	growth	and	decreasing	 inequality,	leading	to	demands	for	better	public	services,	including	improvements	in	healthcare.[4]		
For	many	health	indicators,	the	country	has	made	improvements:	life	expectancy	in	Brazil	has	increased	 by	nearly	20	years	in	the	past	4	decades[5]	and	maternal	mortality	has	nearly	been	cut	in	half.[6]	The	 government’s	health	scheme,	Sistema	Único	de	Saúde	(SUS)	or	the	Unified	Health	System,	is	credited	for	 many	of	these	improvements	as	the	largest	public	healthcare	system	in	the	world,	guaranteeing	 healthcare	to	100%	of	the	Brazilian	population.[7]	It	is	federally	funded	with	additional	tax	revenues	and	 social	contributions	from	the	individual	states	and	municipalities.	The	financing	for	SUS	is	largely	 governed	at	the	federal	level,	though	healthcare	administration	and	management	is	left	to	states	and	 municipalities.[7]	Unfortunately,	SUS	is	vulnerable	to	the	geographic	and	economic	disparities	that	 beleaguer	the	country.	For	example,	65·9%	of	doctors	in	Brazil	practice	in	the	state	capitals	while	only	 24%	of	the	national	population	lives	there.[8]	Similarly,	only	56%	of	physicians	practicing	in	the	North	 completed	residency	training	compared	to	78%	of	physicians	practicing	in	the	Southeast.[8]	
This	maldistribution	of	services	is	seen	in	the	surgical	disciplines	as	well.	Surgical	residency	began	in	 Southeastern	cities	in	1948,	and	even	in	2008,	73%	of	general	surgery	training	positions	remained	in	the	 South	and	Southeast.[9]	While	disparity	in	availability	and	quality	of	surgical	care	is	widely	 acknowledged,[8-10]	it	has	not	yet	been	systematically	studied.		
The	primary	goal	of	this	study	is	to	characterize	the	Brazilian	surgical	health	system	according	to	the	six	 indicators	proposed	by	LCoGS.	In	doing	so,	we	will	provide	an	in-depth	assessment	of	Brazil’s	surgical	 system	while	also	identifying	and	quantifying	geographic	disparities	that	exist.	We	also	use	Brazil	to	gain	 further	understanding	of	the	complexity	of	these	newly	proposed	indicators.		
5	 	

	 METHODS:	
Data	were	obtained	from	national,	open-access	databases	that	are	organized	and	maintained	by	the	 government.	DATASUS	is	the	online	portal	for	an	administrative	database	SIH/SUS	 (http://www2.datasus.gov.br).	SIH/SUS	categorizes	all	SUS	payments	for	hospitalizations	and	 procedures	and	has	been	used	in	numerous	studies.[11]		Instituto	Brasileiro	de	Geografia	e	Estatistica	 (IBGE)	is	the	national	geographic	and	statistical	database	(http://www.ibge.gov.br).	We	accessed	all	 databases	in	September	2015.	
Indicator	1:	Two-Hour	Access	to	Surgery	
In	the	Lancet	Commission	on	Global	Surgery,	two-hour	access	to	surgery	is	defined	as	the	proportion	of	 the	population	that	can	access	a	facility	that	can	perform	caesarean	delivery,	laparotomy	and	open	 fracture	repairs	within	two	hours.[1]	Because	facility-level	data	on	these	procedures	was	unavailable,	we	 queried	the	Cadastro	Nacional	de	Establecimentos	de	Saúde	(CNES),	a	portal	within	DATASUS,	for	public	 hospitals	that	employ	at	least	one	surgeon,	anesthesiologist	and	obstetrician,	and	have	a	doctor	on	call	 24	hours	per	day.	As	such	we	were	unable	to	calculate	a	true	indicator	1	and	the	provided	list	is	simply	a	 proxy	measure.	
Using	proprietary	software	from	Redivis	(https://www.redivis.com),	the	geographic	location	of	these	 hospitals	was	mapped	and	two-hour	distance	was	calculated.	Hospital	location	was	entered	using	 latitude	and	longitude	and	two-hour	distance	was	estimated	using	known	speed	limits	for	roads	and	 highways	and	average	walking	speed	where	no	roads	were	present.	Finally,	the	geographic	area	of	the	 map	generated	to	simulate	two-hour	access	was	compared	against	the	population	density,	from	 WorldPop	(www.worldpop.org.uk),	in	order	to	calculate	the	percentage	of	the	population	with	two-hour	 access	to	an	included	facility.	
Indicator	2:	Surgical	Workforce	Density	
Surgeons,	anesthesiologists	and	obstetricians	(SAO)	were	identified	from	CNES,	which	includes	a	count	 of	all	professionals	registered	in	a	municipality,	for	2014.	“Surgeons”	included	general	surgeons,	surgical	 oncologists,	cardiovascular	surgeons,	orthopedic	surgeons,	hand	surgeons,	otolaryngologists,	head	and	 neck	surgeons,	pediatric	surgeons,	plastic	surgeons,	thoracic	surgeons,	vascular	surgeons,	 ophthalmologists,	urologists	and	neurosurgeons.	“Anesthesiologist”	and	“Obstetricians”	were	 categorized	as	such	in	DATASUS.	A	population	estimate	was	obtained	from	IBGE	and	is	based	on	census	 data	and	projected	growth.	The	surgical	workforce	density	that	we	reported	was	the	total	number	of	 SAO	per	100,000	population	and	was	calculated	separately	for	each	state,	region,	and	for	the	country	as	 a	whole.		
Indicator	3:	Surgical	Volume	
Surgical	volume	was	identified	using	annual	numbers	registered	in	DATASUS	for	the	year	2014.	We	 included	both	inpatient	and	ambulatory	surgeries	in	these	fields:	endocrine,	peripheral	and	central	 nervous	system,	head	and	neck,	ophthalmologic,	vascular,	gastrointestinal	and	abdominal,	orthopedic,	
6	 	

genitourinary,	breast,	obstetric	and	gynecologic,	thoracic,	reconstructive,	and	oncologic.	Exclusion	 criteria	included	anesthetic	and	dental	procedures.		
Indicator	4:	Perioperative	Mortality	Rate	
DATASUS	reports	mortality	data	on	any	deaths	occurring	during	an	inpatient	hospitalization.	 Perioperative	mortality	rate	(POMR)	was	calculated	using	inpatient	mortality	that	occurred	during	a	 hospitalization	in	which	a	patient	underwent	a	procedure.		
Indicator	5:	Protection	against	Impoverishing	Expenditure	
This	indicator	is	defined	as	the	proportion	of	households	protected	against	impoverishment	from	out-ofpocket	payment	for	surgical	and	anesthetic	care.	To	determine	impoverishment,	we	used	the	nationally	 determined	poverty	line.[12]	The	threshold	for	impoverishing	expenditure	would	be	any	income	below:	
������������������������������������������	������������������������	������������	������������������������������������������∙%	������������������	������������	������������������������������������	������������������������������������������������������������������+[������������������������������������������	������������������������]	
For	the	cost	of	surgery,	we	utilized	average	cost	of	surgery	to	SUS	in	Brazilian	Reals,	from	2008	to	2014,	 as	reported	in	DATASUS.	This	value	reflects	the	total	cost	to	SUS	of	an	inpatient	hospitalization	and	 associated	costs	during	which	the	patient	underwent	a	surgical	procedure,	as	defined	above.	The	 expected	proportion	of	out-of-pocket	expenditure	for	healthcare	is	provided	by	the	World	Bank.[13]	A	 well-validated	descriptive	model	of	income	distribution,	the	gamma	distribution,	was	used	to	generate	 income	distributions	for	the	country,	regions,	and	states.[14]	GDP	per	capita	and	the	Gini	Index	were	 used	as	model	parameters.	The	percentage	of	protection	against	impoverishing	expenditure	is	the	 percentage	of	all	incomes	in	the	distribution	that	remain	above	the	poverty	line	after	the	cost	of	surgery	 was	deducted.		
Indicator	6:	Protection	against	Catastrophic	Expenditure	
Methods	for	determining	income	distributions	are	the	same	as	in	Indicator	5.	Catastrophic	expenditure,	 however,	is	defined	as	out-of-pocket	expenditure	on	surgical	and	anesthetic	care	that	is	greater	40%	of	 the	patient’s	post-subsistence	income	where	post-subsistence	income	is	the	household	income	after	 food	expenditure.	Thus,	the	threshold	for	catastrophic	expenditure	is	any	income	below:	
[������������������������������������������	������������������������	������������	������������������������������������������]∙%	������������������	������������	������������������������������������	������������������������������������������������������������������0·40+������������������������������������������	������������������������	������������������������������������������������������������������	
As	in	Indicator	5,	this	income	distribution	was	then	compared	against	the	cost	of	surgery	to	determine	 the	percentage	of	people	protected	against	catastrophic	expenditure.	Food	expenditure	data	is	available	 from	2008	at	the	state	level.	State	food	expenditure	levels	were	modeled	using	the	national	economic	 index	for	years	2009-2014.	
Ethics	Statement	
This	study	was	conducted	according	to	the	principles	expressed	in	the	Declaration	of	Helsinki.	This	study	 was	granted	an	IRB	exemption	by	Boston	Children’s	Hospital	and	an	ethics	exemption	from	the	 Faculdade	de	Medicina	da	Universidade	de	São	Paulo.	The	authors	have	no	conflicts	of	interest	and	 there	was	no	funding	source	for	this	study.	
Statistical	analysis	
7	 	

To	compare	the	regional	variation	in	modeled	parameters	(indicators	5	and	6),	ANOVA	was	performed.	 No	statistical	analysis	was	performed	on	tabulated	data	(indicators	2-4).	All	statistical	analyses	were	 performed	using	statistical	program,	R,	and	Microsoft	Excel.	Geographic	mapping	and	analysis	was	 performed	using	the	geographic	information	system	package,	QGIS	to	visualize	geospatial	information	 generated	or	tabulated	in	R	and	Microsoft	Excel.	 	
8	 	

	 RESULTS:	 Indicator	1:	Two-Hour	Access	to	Surgery	 Our	query	from	the	CNES	yielded	1,247	hospitals	that	met	criteria	nationwide.	We	calculate	a	proxymeasure	for	indicator	1	in	which	97·2%	of	the	population	of	Brazil	has	two-hour	access	to	a	facility	that	 may	be	able	to	provide	surgery.		 Indicator	2:	Surgical	Workforce	Density	 The	surgical	workforce	density	was	34·7	SAO	per	100,000	population	in	2014	(Table	2,	Appendix	1).	 Geographically,	this	ranges	from	18·4	SAO	per	100,000	people	in	the	North	Region,	to	45·81	SAO	per	 100,000	people	in	the	Southeast,	in	2014	(Figure	1a,	Table	2).	Of	the	total	70,449	SAO	in	Brazil,	40,808	 are	surgeons	(58·9%	of	the	surgical	workforce),	11,492	are	anesthesiologists	(16·3%),	and	18,149	are	 obstetricians	(25·8%).	 The	density	of	the	surgical	workforce	that	operates	in	the	public	sector	drops	to	23·0	SAO	per	100,000	 people.	57·9%	of	all	surgeons,	82·7%	of	all	anesthesiologists,	and	63·1%	of	all	obstetricians	maintain	a	 job	in	the	public	sector.		 Indicator	3:	Surgical	Volume	 Surgical	volume	in	the	public	sector	was	4,433	surgeries	per	100,000	people	in	2014.	Geographically,	this	 ranges	from	3,518	surgeries	per	100,000	people	per	year	in	the	North	Region,	to	5,151	surgeries	per	 100,000	people	per	year	in	the	Central	West,	in	2014	(Figure	1b,	Table	2).	 Indicator	4:	Perioperative	Mortality	Rate	 Perioperative	mortality	rate	was	1·71%	in	2014,	ranging	from	1·12%	in	the	North	to	2·13%	in	the	South,	 in	2014	(Figure	1c,	Table	2).	The	mean	inpatient	stay	for	a	surgical	admission	in	Brazil	was	3.6	days	in	 2014.	 The	perioperative	mortality	rate	for	a	cesarean	section	alone	is	0.034%	in	Brazil.	The	rate	is	highest	in	 the	Northeast	(0.037%)	and	lowest	in	the	South	Region	(0.025%).	 Indicator	5:	Protection	against	Impoverishing	Expenditure	 Protection	against	impoverishing	expenditure	was	79·4%	in	2014,	ranging	from	68·7%	in	the	Northeast	 Region,	up	to	82·3%	in	the	Central	West	Region,	in	2014	(Figure	1d,	Table	2).		 Indicator	6:	Protection	against	Catastrophic	Expenditure	 Protection	against	catastrophic	expenditure	was	84·6%	in	2014,	ranging	from	78·5%	in	the	Northeast	 Region,	up	to	87·2%	in	the	Central	West	Region,	in	2014	(Table	2).	
9	 	

	 DISCUSSION:	
We	used	the	Brazilian	public	health	sector	database	to	characterize	Brazil’s	surgical	system	according	to	 six	indicators.	Based	on	limited	international	data,	we	find	that	Brazil	falls	within	the	range	of	values	 seen	by	other	upper-middle	income	countries	and	performs	substantially	better	than	low-income	 countries.[1	15]	Moreover,	we	find	that	Brazil	approaches	or	meets	benchmarks	that	suggest	adequate	 national	workforce	and	surgical	volume,	yet	the	country	suffers	from	problems	of	geographic	 distribution	and	equity.	This	analysis	of	the	surgical	system	is	timely	in	light	of	a	recent	survey	of	the	 Brazilian	public	which	found	that	93%	of	people	have	a	negative	perception	of	Brazilian	healthcare,	with	 access	to	surgery	as	the	largest	perceived	problem.[10]		
Despite	the	existence	of	a	national	health	program,	regional	variation	in	the	surgical	indicators	mirror	 economic	and	health	disparities	in	the	country.	The	North	and	Northeast,	the	two	poorest	regions	of	 Brazil,	have	a	lower	density	of	surgical	providers,	fewer	surgical	procedures	performed	per	capita,	and	 more	patients	are	impoverished	due	to	surgery.	In	some	cases,	the	disparity	is	severe.	In	the	 Northeastern	state	of	Piaui,	nearly	40%	of	patients,	twice	the	national	figure,	are	impoverished	due	to	 surgical	costs.	Additionally,	in	the	Northeastern	state	of	Maranhão,	there	are	eight	times	fewer	 SAO/100,000	people	than	in	the	capital	district	of	Brasilia	(11·4	versus	88·6,	respectively).	
Workforce	shortages	have	long	been	perceived	to	be	a	problem	in	the	Brazilian	health	system.[16]	For	 specialist	surgical	care,	we	found	the	wealthier	regions	meet	benchmarks	for	density	of	surgical	 providers	but	that	poorer	North	and	Northeastern	regions	in	Brazil	barely	meet	the	threshold	of	20-40	 SAO/100,000	that	is	a	suggested	minimum	to	meet	a	region’s	needs.[1	17]		To	address	a	general	 shortage	of	doctors,	in	2013	the	Brazilian	government	launched	the	Mais	Médicos	(More	Doctors)	 scheme.[18]	The	scheme’s	primary	focus	is	on	increasing	the	number	of	primary	care	doctors	and	 strengthening	Primary	Healthcare	Units.[18]	In	addition	to	this,	the	scheme	strives	to	increase	medical	 school	enrollment	outside	of	state	capitals	and	create	12,400	residency	spots	by	2018;	the	goal	of	this	is	 to	ensure	every	Brazilian	medical	graduate	is	able	to	pursue	post-graduate	training	and	that	they	are	 able	to	do	so	closer	to	the	communities	they	come	from.[18]		
While	the	focus	of	this	effort	is	on	increasing	training	for	primary	care	specialties,	a	parallel	effort	in	 surgical	care	is	warranted.	Increasing	surgical	and	anesthesia	training	in	rural	areas	may	be	a	useful	 retention	strategy	for	the	rural	surgical	workforce,	as	has	been	demonstrated	at	a	hospital	in	Kenya.[19]	 The	development	of	a	National	Specialist	Registration,	currently	underway,	is	likely	to	assist	in	 identifying	targets	for	public	policies	on	specialist	training.[18]	For	example,	the	shortage	of	 anesthesiologists	–	the	SAO	workforce	in	brazil	is	comprised	of	16·3%	anesthesiologists,	under	the	 worldwide	median	of	20%[15]	–	suggests	that	reallocating	residency	spots	may	be	useful.	
	Apart	from	geographic	analyses,	in	a	country	such	as	Brazil	where	broad	benchmarks	of	SAO	density	are	 met,	additional	disaggregation	is	also	valuable.	We	find,	for	example,	the	SAO	density	of	public	sector	 providers	is	lower	than	overall	providers	by	nearly	12	per	100,000.	True	availability	is	likely	even	worse	 than	our	numbers	show:	in	Brazil,	over	half	of	all	of	the	physicians	have	three	or	more	job	positions	and	 70·5%	of	all	physicians	in	the	public	sector	also	work	in	the	private	sector.[16]	The	time	distribution	of	 time	each	specialist	spends	in	the	public	versus	private	sector	may	be	heavily	biased	towards	the	private	
10	 	

sector.	Thus	consideration	of	public	full-time	equivalents	may	give	a	more	accurate	picture	of	the	 surgical	workforce	available	to	the	75-80%	of	Brazilians	who	rely	exclusively	on	SUS.[20]	
In	Brazil,	the	national	surgical	volume	of	4,337	procedures	/	100,000	in	the	public	sector	alone	is	close	to	 the	suggested	minimum	of	5,000	procedures	/	100,000.	In	light	of	this,	it	is	important	to	turn	attention	 to	quality	and	case-mix	of	the	surgeries	being	performed	and	the	appropriateness	of	personnel	to	do	 them.	In	particular,	selective	overprovision	of	surgeries	such	as	cesarean	section[21-23]	may	be	harmful	 to	patients	and	draw	resources	away	from	the	provision	of	other	necessary	procedures.		
Additionally,	an	integrated	consideration	of	the	indicators	is	valuable.	We	find,	for	example,	the	state	of	 Rio	de	Janeiro	has	a	high	workforce	density	yet	provides	a	very	low	number	of	procedures	in	the	public	 sector	yielding	a	ratio	of	41.7	procedures	per	SAO	provider.	The	state	of	Pará,	conversely,	has	high	 public	sector	output	with	a	low	workforce	density,	with	a	ratio	of	305.2	procedures	per	SAO	provider.		 These	findings	may	be	driven	by	the	fact	that	surgeons	in	Rio	disproportionately	provide	care	in	the	 private	sector.	Further	investigation	into	this	type	of	variation	is	warranted.	
Our	study	also	found	that	inpatient	POMR	demonstrates	a	counterintuitive	geographic	disparity.	The	 South	and	Southeastern	states,	which	have	a	greater	proportion	of	academic	surgical	programs,[9]	 suffer	a	considerably	higher	POMR	than	the	Northern	and	Northeastern	states.	This	may,	in	part,	be	 driven	by	different	patient	populations,	different	physician	behavior,	or	underreporting	of	mortality.	In	 the	South	and	Southeast,	more	highly	trained	surgeons	with	greater	resources	may	perform	highercomplexity	procedures	on	patients	with	greater	comorbidities,	thereby	driving	up	POMR.	POMR	for	 cesarean	section	alone,	a	procedure	for	which	there	is	less	variation	of	risk,	demonstrates	the	expected	 findings	of	lower	POMR	in	the	South	and	Southeast	than	in	the	North	and	Northeast.	An	additional	 factor	may	be	underreporting	of	mortality,	which	is	known	to	occur	with	greater	frequency	in	the	North	 and	Northeast	of	Brazil	than	in	other	regions.[24]	
It	is	additionally	worth	noting	that	our	definition	of	POMR	evaluated	mortality	over	the	duration	of	the	 inpatient	hospitalization.	Recent	studies	of	POMR	in	Brazil,	have	reported	substantially	lower	mortality	 rates	ranging	from	0.16%	to	0.51%	in	tertiary	hospitals	in	São	Paulo.[25	26]	These	studies,	however,	 have	been	single-center	studies	with	a	follow-up	period	no	greater	than	24	hours	from	procedure.	The	 POMR	we	report	is	provided	over	a	much	longer	duration	of	follow-up.		
Worldwide,	33	million	people	each	year	face	catastrophic	expenditures	due	to	surgical	costs.[1	27]	SUS	 strives	to	mitigate	this	for	Brazilians	by	eliminating	user	fees.	The	World	Bank,	however,	estimates	that	 29·9%	of	public	health	expenses	in	Brazil	are	passed	on	to	the	patient	as	out-of-pocket	expenditure.[13],	 While	studies	have	reported	much	of	this	expenditure	is	on	non-procedural	costs	such	as	medicines,[28	 29]	it	is	likely	that	the	associated	expenses	of	surgery	exert	a	substantial	burden	on	low-income	families.	 Non-medical	costs	such	as	transportation	–	which	were	not	accounted	for	in	our	study	–	can	also	 amount	to	an	enormous	economic	burden.[30]		As	SUS	strives	to	truly	achieve	universally	affordable	 surgical	care,	consideration	of	these	out-of-pocket	expenses	and	how	they	may	vary	by	region	is	critical.	 To	do	so,	will	require	an	understanding	of	patient	perspectives	on	financing	for	healthcare.		
This	study	has	limitations.	Notably,	we	were	not	able	to	calculate	two-hour	access	to	a	surgical	facility	 due	to	the	fact	that	we	could	not	accurately	identify	which	hospitals	are	currently	equipped	with	
11	 	

operating	rooms,	which	hospitals	have	on-call	anesthesiologists,	and	which	hospitals	have	physicians	 capable	of	the	performing	cesarean	section,	management	of	open	fractures,	and	laparotomy.	Because	 of	this,	the	proxy	measure	we	report	likely	overestimated	the	population’s	true	two-hour	access	to	 surgical	care.	Furthermore,	our	estimation	assumes	access	to	a	motor	vehicle,	no	road	traffic,	and	the	 presence	of	a	functional	operating	room	at	the	hospital.	A	true	measurement	of	two-hour	geographic	 access	would	require	facility-level	analysis	of	hospital	surgical	capabilities,	workforce	availability	with	 full-time	equivalents,	and	patient	transportation	methods.	In	order	to	carry	this	out	on	a	population	 level,	a	sampling	strategy	would	likely	need	to	be	used	and	extrapolated.	Further	studies	exploring	this	 particular	indicator	are	warranted.	Moreover,	this	indicator	could	be	buttressed	by	reporting	operating	 theater	density	/	100,000.		 An	additional	limitation	of	our	data	is	the	fidelity	of	our	data	sources.	Reporting	concerns	have	been	 raised	about	admission	diagnoses	in	DATASUS,	though	this	is	a	parameter	we	did	not	use	in	our	 analysis.[11]	In	addition	to	variable	mortality	reporting,	misreporting	of	employment	is	thought	to	be	 prevalent,	as	DATASUS	collects	data	on	job	titles	but	not	necessarily	on	actual	specialist	training.	While	 we	are	able	to	estimate	indicator	2	from	this	data,	it	is	limited	by	the	dataset	it	is	drawn	from.	Similarly,	 out	of	pocket	spending	on	healthcare	likely	differs	by	income	level[28]	which	could	lead	to	our	reported	 numbers	for	indicators	5	and	6	underestimating	the	true	value.	 A	final	limitation	of	our	study	is	its	focus	on	the	public	sector.	While	SUS	is	available	to	all,	there	remains	 a	complex	private-public	mix	for	healthcare	provision.[5	31]	This	public	spending	on	healthcare	accounts	 for	4.7%	of	GDP[32]	while	private	spending	is	thought	to	account	for	an	additional	5·0%	of	GDP.[33]	 Private	insurance	covers	between	20-25%	of	the	population[20]	and	private	facilities	frequently	care	for	 both	SUS	and	private	patients.	Nevertheless,	our	findings	apply	to	the	system	that	cares	for	the	vast	 majority	of	Brazilians.		 Despite	these	limitations,	this	study	is	able	to	utilize	a	detailed	and	publicly	available	dataset	along	with	 robust	modelling	techniques	to	provide	a	systematic	analysis	of	Brazil’s	surgical	system.	We	evaluate	 areas	of	surgical	preparedness,	care	delivery,	and	financial	impact	on	patients.	Our	analysis	is	unique	in	 its	evaluation	of	surgical	care	at	the	state	level	and	provides	insight	into	regional	variability.	From	our	 assessment	we	are	able	to	identify	areas	in	which	Brazil’s	surgical	system	can	be	improved	while	also	 gaining	insight	into	these	newly	proposed	surgical	indicators.	These	findings	are	summarized	in	Table	3.		 	 	
12	 	

	 CONCLUSIONS:	 Brazil	is	a	vast	and	populous	country	with	a	public	surgical	system	that	should	be	applauded	for	meeting	 several	key	benchmarks.	Nevertheless,	dissatisfaction	with	surgical	care	in	Brazil	remains	high.[9]	Our	 analysis	points	to	large	geographic	disparities	which	raise	concerns	of	equity.	Policies	should	focus	on	 stimulating	appropriate	geographic	allocation	of	the	surgical	workforce,	addressing	the	deficits	of	 certain	specialists,	and	better	distributing	surgical	volume.	An	integrated	and	in-depth	consideration	of	 these	indicators	will	uncover	various	inequities	in	care	provision	that	must	be	further	investigated	and	 addressed.		This	may	involve	patient-	and	facility-level	analyses	to	better	understand	key	issues	related	 to	geographic	and	financial	access	to	car	or	careful	consideration	of	quality	of	care	when	other	indicator	 benchmarks	are	met.	 This	study	also	demonstrates	that	the	collection	of	surgical	system	indicators	is	possible	where	national	 health	data	is	recorded	and	made	available.	This	type	of	analysis	can	be	encouraged	for	all	nations	 seeking	to	understand	their	surgical	systems.	 	
13	 	

	
REFERENCES:	
1.	Meara	JG,	Leather	AJ,	Hagander	L,	et	al.	Global	Surgery	2030:	evidence	and	solutions	for	achieving	 health,	welfare,	and	economic	development.	Lancet	2015;386(9993):569-624	doi:	 10.1016/S0140-6736(15)60160-X[published	Online	First:	Epub	Date]|.	
2.	Ng-Kamstra	JS,	Greenberg	SLM,	Abdullah	F,	et	al.	Global	Surgery	2030:	a	roadmap	for	high	income	 country	actors.	BMJ	Global	Health	2016;1(1)		
3.	Victora	CG,	Barreto	ML,	do	Carmo	Leal	M,	et	al.	Health	conditions	and	health-policy	innovations	in	 Brazil:	the	way	forward.	Lancet	2011;377(9782):2042-53	doi:	10.1016/S0140-6736(11)60055X[published	Online	First:	Epub	Date]|.	
4.	Abrucio	FL.	Recent	trajectory	of	the	Brazilian	public	management:	a	critical	assessment	and	the	 renewal	of	the	reform	agenda.	Revista	de	Administração	Pública	2007;41:67-86		
5.	Paim	J,	Travassos	C,	Almeida	C,	et	al.	The	Brazilian	health	system:	history,	advances,	and	challenges.	 Lancet	2011;377(9779):1778-97	doi:	10.1016/S0140-6736(11)60054-8[published	Online	First:	 Epub	Date]|.	
6.	Global	Health	Observatory	data	repository:	Maternal	mortality	ratio:	World	Health	Organization,	 2013.	
7.	Marten	R,	McIntyre	D,	Travassos	C,	et	al.	An	assessment	of	progress	towards	universal	health	 coverage	in	Brazil,	Russia,	India,	China,	and	South	Africa	(BRICS).	Lancet	2014;384(9960):2164-71	 doi:	10.1016/S0140-6736(14)60075-1[published	Online	First:	Epub	Date]|.	
8.	Birolini	D,	Ferreira	EA,	Rasslan	S,	et	al.	Surgery	in	Brazil.	Archives	of	surgery	2002;137(3):352-8		 9.	Gomes	dos	Santos	E.	General	surgery	residency	in	Brasil	-	very	far	from	real	practice.	Revista	do	
Colégio	Brasileiro	de	Cirurgiões	2009;36(3):271-76		 10.	93%	têm	percepção	negativa	da	saúde	Jornal	do	Conselho	Federal	de	Medicina	2015;249:10		 11.	Bittencourt	SA,	Camacho	LAB,	Leal	MdC.	O	Sistema	de	Informação	Hospitalar	e	sua	aplicação	na	
saúde	coletiva.	Cadernos	de	Saúde	Pública	2006;22(1)		 12.	Bolsa	Família.	Secondary	Bolsa	Família		2015.	http://www.caixa.gov.br/programas-sociais/bolsa-
familia/Paginas/default.aspx.	 13.	Out-of-pocket	health	expenditure	(%	of	total	expenditure	on	health).	World	Development	Indicators:	
World	Bank,	2013.	 14.	Salem	ABZ,	Mount	TD.	Convenient	Descriptive	Model	of	Income-Distribution	-	Gamma	Density.	
Econometrica	1974;42(6):1115-27	doi:	Doi	10.2307/1914221[published	Online	First:	Epub	 Date]|.	 15.	Ng-Kamstra	JS,	Raykar	NP,	Lin	Y,	et	al.	Data	for	the	sustainable	development	of	surgical	systems:	A	 global	collaboration.	2015	December	2015.	 http://media.wix.com/ugd/346076_26d4a927dc074b128be2e61f34e14018.pdf	(accessed	 2/2/2016).	 16.	Scheffer	M.	DEMOGRAFIA	MÉDICA	NO	BRASIL.	São	Paulo:	Conselho	Federal	de	Medicina,	2011-2015.	 17.	Holmer	H,	Shrime	MG,	Riesel	JN,	et	al.	Towards	closing	the	gap	of	the	global	surgeon,	 anaesthesiologist,	and	obstetrician	workforce:	thresholds	and	projections	towards	2030.	Lancet	 2015;385	Suppl	2:S40	doi:	10.1016/S0140-6736(15)60835-2[published	Online	First:	Epub	Date]|.	 18.	MAIS	MÉDICOS	Conheça	o	Programa.	Secondary	MAIS	MÉDICOS	Conheça	o	Programa.	 http://www.maismedicos.gov.br/conheca-programa.	 19.	Newton	M,	Bird	P.	Impact	of	parallel	anesthesia	and	surgical	provider	training	in	sub-Saharan	Africa:	 a	model	for	a	resource-poor	setting.	World	journal	of	surgery	2010;34(3):445-52	doi:	 10.1007/s00268-009-0195-z[published	Online	First:	Epub	Date]|.	
14	 	

20.	Dados	do	setor	de	saúde	suplementar.	Secondary	Dados	do	setor	de	saúde	suplementar.	 http://www.ans.gov.br/aans/noticias-ans/numeros-do-setor/3030-dados-do-setor-de-saudesuplementar.	
21.	Barros	FC,	Matijasevich	A,	Maranhao	AG,	et	al.	Cesarean	sections	in	Brazil:	will	they	ever	stop	 increasing?	Revista	panamericana	de	salud	publica	=	Pan	American	journal	of	public	health	 2015;38(3):217-25		
22.	Gibbons	L,	Belizán	JM,	Lauer	JA,	et	al.	The	Global	Numbers	and	Costs	of	Additionally	Needed	and	 Unnecessary	Caesarean	Sections	Performed	per	Year:	Overuse	as	a	Barrier	to	Universal	 Coverage:	World	Health	Organization,	2010.	
23.	Wise	J.	Target	rates	for	caesarean	section	may	be	too	low,	say	researchers.	Bmj	2015;351:h6439	doi:	 10.1136/bmj.h6439[published	Online	First:	Epub	Date]|.	
24.	Razão	entre	óbitos	informados	e	estimados	-	F.11	-	2006.	Secondary	Razão	entre	óbitos	informados	 e	estimados	-	F.11	-	2006		10/1/2015	2006.	http://fichas.ripsa.org.br/2006/F-11/?l=en_US.	
25.	Braz	LG,	Modolo	NS,	do	Nascimento	P,	Jr.,	et	al.	Perioperative	cardiac	arrest:	a	study	of	53,718	 anaesthetics	over	9	yr	from	a	Brazilian	teaching	hospital.	British	journal	of	anaesthesia	 2006;96(5):569-75	doi:	10.1093/bja/ael065[published	Online	First:	Epub	Date]|.	
26.	Pignaton	W,	Braz	JR,	Kusano	PS,	et	al.	Perioperative	and	Anesthesia-Related	Mortality:	An	8-Year	 Observational	Survey	From	a	Tertiary	Teaching	Hospital.	Medicine	2016;95(2):e2208	doi:	 10.1097/MD.0000000000002208[published	Online	First:	Epub	Date]|.	
27.	Shrime	MG,	Dare	AJ,	Alkire	BC,	et	al.	Catastrophic	expenditure	to	pay	for	surgery	worldwide:	a	 modelling	study.	The	Lancet	Global	health	2015;3	Suppl	2:S38-44	doi:	10.1016/S2214109X(15)70085-9[published	Online	First:	Epub	Date]|.	
28.	Barros	AJ,	Bertoldi	AD.	Out-of-pocket	health	expenditure	in	a	population	covered	by	the	Family	 Health	Program	in	Brazil.	International	journal	of	epidemiology	2008;37(4):758-65	doi:	 10.1093/ije/dyn063[published	Online	First:	Epub	Date]|.	
29.	Silveira	FG,	Osório	RG,	Piola	SF.	Os	gastos	das	famílias	com	saúde.	Ciência	&	Saúde	Coletiva	2002;7(4)		 30.	Nahar	S,	Costello	A.	The	hidden	cost	of	'free'	maternity	care	in	Dhaka,	Bangladesh.	Health	policy	and	
planning	1998;13(4):417-22		 31.	Victora	CG,	Aquino	EM,	do	Carmo	Leal	M,	et	al.	Maternal	and	child	health	in	Brazil:	progress	and	
challenges.	Lancet	2011;377(9780):1863-76	doi:	10.1016/S0140-6736(11)60138-4[published	 Online	First:	Epub	Date]|.	 32.	Health	expenditure,	public	(%	of	GDP).	World	Development	Indicators:	World	Bank,	2013.	 33.	Health	expenditure,	private	(%	of	GDP).	World	Development	Indicators:	World	Bank,	2013.	
	
15	 	

	 TABLE	AND	FIGURE	LEGENDS:	 Table	1:	Six	Lancet	Indicators	for	measurement	and	assessment	of	global	surgical	systems.	 Table	2:	Indicators	2	through	6	for	each	region	in	Brazil,	in	the	year	2014.	SAO:	surgeons,	 anesthesiologists,	and	obstetricians;	Vol:	Surgical	Volume;	POMR:	Perioperative	Mortality	Rate;	C/S:	 Caesarean	Section;	CI:	confidence	Interval;	*One	Way	ANOVA	performed	between	regions	 Table	3:	Recommendations	for	surgical	system	strengthening	in	Brazil	and	globally	for	the	Lancet	 Indicators.	 Figure	1a:	Indicator	2:	Total	surgeon,	anesthesiologist,	and	obstetrician	workforce	density	per	100,000	 people,	by	state	in	the	year	2014.	 Figure	1b:	Indicator	3:	Total	surgical	volume	per	100,000	people,	by	state	in	the	year	2014.	 Figure	1c:	Indicator	4:	Perioperative	mortality	rate	by	state,	in	the	year	2014.	 Figure	1d:	Indicator	5:	Protection	against	impoverishing	expenditure	by	state,	in	the	year	2014.	 Appendix	1:	Indicators	2	through	6	for	each	region	and	state	in	Brazil,	in	the	year	2014,	with	additional	 sub-indicators	for	physicians	working	in	the	public	sector	and	cesarean	section	perioperative	mortality	 rate.	SAO:	Surgeons,	Anesthesiologists,	and	Obstetricians;	Vol:	Surgical	Volume;	POMR:	Perioperative	 Mortality	Rate;	CS:	Cesarean	Section;	CI:	Confidence	Interval;	*One	Way	ANOVA	performed	between	 regions		
16	 	

	 Table	1:	Six	Lancet	Indicators	for	measurement	and	assessment	of	global	surgical	systems.	

Group	1:	Preparedness	for	surgical	and	anesthesia	care	

Access	to	timely	essential	 Proportion	of	the	population	that	can	

surgery	

access,	within	2	hours,	a	facility	that	can	

do	caesarean	delivery,	laparotomy,	and	

treatment	of	open	fracture	(the	

Bellwether	Procedures)	

Specialist	surgical	

Number	of	specialist	surgical,	

workforce	density	

anesthetic,	and	obstetric	physicians	

who	are	working	per	100	000	

population	

Group	2:	Delivery	of	surgical	and	anesthesia	care	

Surgical	volume	

Procedures	done	in	an	operating	

theatre,	per	100	000	population	per	

year	

Perioperative	mortality	 All-cause	death	rate	before	discharge	in	

rate	 patients	who	have	had	a	procedure	in	

an	operating	theatre,	divided	by	the	

total	number	of	procedures,	presented	

as	a	percentage	

Group	3:	Effect	of	surgical	and	anesthesia	care	

Protection	against	

Proportion	of	households	protected	

impoverishing	

against	impoverishment	from	direct	

expenditure	

out-of-pocket	payments	for	surgical	and	

anesthesia	care	

Protection	against	

Proportion	of	households	protected	

catastrophic	expenditure	 against	catastrophic	expenditure	from	

direct	out-of-pocket	payments	for	

surgical	and	anesthesia	care	

Adapted	from	The	Lancet	Commission	on	Global	Surgery	

Targets	 A	minimum	of	80%	coverage	 of	essential	surgical	and	 anaesthesia	services	per	 country	by	2030	
100%	of	countries	with	at	 least	20	surgical,	anaesthetic,	 and	obstetric	physicians	per	 100	000	population	by	2030	 	 Minimum	of	5000	procedures	 per	100	000	population	by	 2030	 No	target	set.	Will	be	reevaluated	upon	further	data	 collection	
	 100%	protection	against	 impoverishment	from	out-ofpocket	payments	for	surgical	 and	anaesthesia	care	 100%	protection	against	 catastrophic	expenditure	from	 out-of-pocket	payments	for	 surgical	and	anaesthesia	care	 by	2030	

17	 	

	 Table	2:	Indicators	2	through	6	for	each	region	in	Brazil,	in	the	year	2014.	

Indicator	2	 Indicator	3	 Indicator	4	 Indicator	5	

Indicator	6	

Region/State	 North	Region	 Northeast	Region	 Southeast	Region	 South	Region	 Central	West	Region	 Brazil	

SAO/100,000	 Vol/100·000	

18·42	

3518·58	

23·59	

4190·93	

45·81	

4742·72	

30·76	

4163·59	

40·03	

5151·20	

34·74	

4433·44	

POMR	 1·12	 1·38	 1·87	 2·13	 1·55	 1·71	

mean	(95%	CI)	 73·85	(73·85-
73·85)	 68·7	(68·7-
68·7)	 82·64	(82·64-
82·64)	 81·39	(81·39-
81·39)	 82·27	(82·27-
82·27)	 79·39	(79·39-
79·39)	

mean	(95%	CI)	 82·24	(82·23-
82·24)	 78·49	(78·49-
78·49)	 86·59	(86·59-
86·59)	 84·77	(84·77-
84·78)	 87·15	(87·15-
87·15)	 84·58	(84·58-
84·59)	

p	Value*	

---	 --- --- <0·01	 <0·01	

SAO:	Surgeons,	Anesthesiologists,	and	Obstetricians;	Vol:	Surgical	Volume;	POMR:	Perioperative	Mortality	

Rate;	CI:	Confidence	Interval;	*One	Way	ANOVA	performed	between	regions	

	

18	 	

	

Table	3:	Recommendations	for	surgical	system	strengthening	in	Brazil	and	globally	for	the	Lancet	

Indicators.	

Surgical	System	Strengthening	

Global	Application	

Indicator	

in	Brazil	

of	the	Lancet	Indicators	

Indicator	1:		 Access	to	timely	 essential	surgery	
Indicator	2:		 Specialist	surgical	 workforce	density	
Indicator	3:		 Surgical	volume	
Indicator	4:		 Perioperative	 mortality	rate	 Indicator	5	and	6:		 Protection	against	 impoverishing	and	 catastrophic	 expenditure	

• Facility-level	data	is	needed	 	
• Address	large	geographic	 disparities	for	SAO	
• Consider	rural	residency	 training	in	addition	to	rural	 medical	education	
• Address	the	shortage	of	 anesthesia	providers	
• Measure	and	report	private	 surgical	volume	
• Address	geographic	 disparities	in	public-sector	 volume	
• Monitor	overuse	as	access	 continues	to	improve	
• Ensure	accurate	reporting	 • Develop	a	national	strategy	
for	assessing	and	improving	 postoperative	outcomes	
• Investigate	what	 expenditures	are	being	 passed	on	to	the	patient	
• Expand	financial	riskprotection	beyond	the	cost	 of	procedures	

• Facility	level	data	is	needed	 • Look	at	operating	room	
density/100,000	population	as	 adjunct	indicator	when	basic	 parameters	of	access	are	met	 • Disaggregate	the	surgical	 workforce	density	by	specialty	 to	find	nuances	 • Assess	internal	distribution	to	 look	for	regional	deficiencies	 • Use	full-time	equivalent	of	 SAO	in	the	public	sector	as	an	 adjunct	indicator	 • Assess	internal	distribution	to	 look	for	regional	deficiencies	 • Monitor	overuse	by	evaluating	 case-mix	
• Consider	procedure-specific	 perioperative	mortality	rates	 to	minimize	variation	in	 patient-risk	
• As	an	adjunct,	consider	 patient-level	analyses	to	 disaggregate	what	contributes	 to	out-of-pocket	expenditure	

	 	

	

	

19