Global, regional, and national age-sex specific mortality for 264 causes of death, 1980–2016: a systematic analysis for the Global Burden of Disease Study 2016

May combine two claims.

Most countries with a 5-star time-series rating were high-SDI, but some high-SDI countries such as France, Poland, and Puerto Rico were rated 4 stars.

“While most countries with a 5-star time series rating were high-SDI countries, some high-SDI countries were rated at 4 stars, such as France, Poland, and Puerto Rico.”

May combine two claims.

Years of life lost were notably lower than expected for neonatal preterm birth complications in many locations in south Asia and southeast Asia, and for cerebrovascular disease in western Europe.

“Other leading causes of YLLs for which YLLs were notably lower than expected included neonatal preterm birth complications in many locations in both south Asia and southeast Asia, and cerebrovascular disease in western Europe.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

May combine two claims.

Communicable diseases declined as a leading source of death, and reductions in HIV/AIDS, malaria, tuberculosis, and diarrhoeal diseases drove much of that decline.

“Generally, communicable diseases decreased as a leading source of death, and much of this decrease was driven by reductions in large contributors to global mortality, including HIV/AIDS, malaria, tuberculosis, and diarrhoeal diseases (table 2).”

Probably not a substantial claim: may not pass S3 (load-bearing).

The GBD 2016 version of the Socio-demographic Index correlates with the GBD 2015 version at 0.977 (p<0.0001).

“the correlation of the GBD 2015 and GBD 2016 versions of SDI is 0·977 (p<0·0001)”

• cites [21] 2017, Global, regional, and national under-5 mortality, adult mortality, age-specific mortality, and life expectancy, 1970–2016: a systematic analysis for the Global Burden of Disease Study 2016, Lancet sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Between 1990 and 2016, total deaths declined in age groups 0-6 days, 7-27 days, 28-364 days, 1-4 years, 5-9 years, 10-14 years, 15-19 years, and 20-24 years, and increased by more than 60% in age groups 80-84, 85-89, 90-94, and 95 years and older.

“Total deaths declined in the age group intervals of 0–6 days, 7–27 days, 28–364 days, 1–4 years, 5–9 years, 10–14 years, 15–19 years, and 20–24 years, and increased by more than 60% in age groups 80–84 years, 85–89 years, 90–94 years, and 95 years and older.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

May combine two claims.

Tuberculosis caused an estimated 1.21 million deaths in 2016, a 20.9% decrease since 2006.

“An estimated 1·21 million deaths (95% UI 1·16 million to 1·27 million) were caused by tuberculosis in 2016, a decrease of 20·9% (17·9–24·5) since 2006.”

May combine two claims.

From 2006 to 2016, the ten leading causes by age-standardised years of life lost rate fell with a median annualised change of −2.89%, while all other causes had a median annualised change of −1.59%.

“The ten leading causes by rates of age-standardised YLL significantly decreased from 2006 to 2016 (median annualised rate of change was a decrease of 2·89%); the median annualised rate of change for all other causes was lower (a decrease of 1·59%) during the same interval.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

May combine two claims.

In 2016, 25 countries had 5-star data quality, 48 had 4 stars, 30 had 3 stars, 21 had 2 stars, and 44 had 1 star.

“Using this rating system, 25 countries were assigned 5 stars, 48 countries had 4 stars, 30 countries had 3 stars, 21 countries had 2 stars, and 44 countries were assigned 1 star (figure 2).”

May combine two claims.

Years of life lost increased globally for causes such as diabetes mellitus and some neoplasms, and in some locations for causes such as drug use disorders and conflict and terrorism.

“YLLs have increased globally for causes such as diabetes mellitus or some neoplasms, and in some locations for causes such as drug use disorders, and conflict and terrorism.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

Probably not a substantial claim: may not pass S3 (load-bearing).

Some high-middle-SDI countries such as Argentina, Brazil, and Israel received 4-star data quality ratings.

“Some high-middle-SDI countries such as Argentina, Brazil, and Israel also received data quality ratings of 4 stars.”

May combine two claims.

From 2006 to 2016, total deaths from communicable, maternal, neonatal, and nutritional diseases fell by 23.9% (95% UI 21.6-26.1), and their age-standardised death rate fell by 32.3% (30.3-34.2).

“total CMNN deaths decreased by 23·9% (95% UI 21·6–26·1), while age-standardised death rates decreased by 32·3% (30·3–34·2).”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

May combine two claims.

For people older than 70, causes of death that increased by more than 90% between 1990 and 2016 included neurological disorders; diabetes, urogenital, blood, and endocrine diseases; unintentional injuries; other non-communicable diseases; musculoskeletal disorders; and mental and substance use disorders.

“Causes of deaths for those aged older than age 70 years that increased by more than 90% were neurological disorders; diabetes, urogenital, blood, and endocrine diseases; unintentional injuries; other non-communicable diseases; musculoskeletal disorders; and mental and substance use disorders.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

May combine two claims.

Between 1990 and 2016, deaths at ages 90-94 increased by 178% (95% UI 176-181) and deaths at ages 95 years and older increased by 210% (208-212).

“Between 1990 and 2016, a profound shift toward deaths at older ages occurred with a 178% (95% UI 176–181) increase in deaths in ages 90–94 years and a 210% (208–212) increase in deaths older than age 95 years.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

May combine two claims.

From 2006 to 2016, the global age-standardised death rate from non-communicable diseases fell by 12.1% (95% UI 10.6-13.4), reaching 614.1 deaths per 100,000 in 2016.

“the global age-standardised NCD death rate decreased 12·1% (10·6–13·4), to 614·1 deaths (603·0–625·3) per 100 000 in 2016.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

May combine two claims.

Increases greater than 30% occurred for prostate cancer, pancreatic cancer, and other neoplasms from 2006 to 2016.

“From 2006 to 2016, increases of greater than 30% occurred for several neoplasms that were large causes of deaths (greater than 200 000 in 2016): prostate cancer (30·8% [95% UI 24·5–36·6], to 381 000 deaths [321 000–413 000]); pancreatic cancer (30·2% [26·2–33·7], to 405 000 deaths [394 000–416 000]); and other neoplasms (30·0% [24·5–33·0], to 431 000 deaths [393 000–444 000]).”

May combine two claims.

The largest decreases in injury deaths from 2006 to 2016 were for exposure to forces of nature and drowning, while the largest increase was for conflict and terrorism.

“The largest decreases in injury deaths from 2006 to 2016 occurred for exposure to forces of nature (49·2% [95% UI 35·2–63·2], to 7060 deaths [4220–10 100]) and drowning (19·0% [13·3–22·2], to 303 000 deaths [273 000–322 000]), while the largest increase was for conflict and terrorism at 143·3% (42·6–370·6), from 61,900 deaths (33 100 to 91 000) deaths to 150 000 deaths [101 400–202 700]).”

May combine two claims.

Dengue was the only neglected tropical disease with a significant increase in cause-specific mortality, with total deaths rising 81.8% from 20 800 in 2006 to 37 800 in 2016.

“Dengue was the only NTD with a significant increase in cause-specific mortality, with an 81·8% (95% UI 42·3–132·6) increase in total deaths, from 20 800 deaths (6000–26 500) in 2006 to 37 800 (10 900–52 700) deaths in 2016”

May combine two claims.

Nutritional deficiencies accounted for 3.49% of total CMNN deaths, or 368 100 deaths in 2016.

“Deaths from nutritional deficiencies constituted 3·49% (95% UI 3·31–3·79) of total deaths due to CMNN causes, resulting in 368 100 (334 000–422 700) in 2016.”

May combine two claims.

Post-neonatal deaths decreased 34.5% to 1.49 million, and deaths among children aged 1-4 years decreased 37.2% to 1.35 million between 2006 and 2016.

“Total deaths among post-neonates (aged 28–364 days) and children aged 1–4 years decreased between 2006 and 2016 by 34·5% (95% UI 31·3–37·4) to 1·49 million deaths (1·42 million to 1·55 million) and 37·2% (33·3–40·5) to 1·35 million deaths (1·28 million to 1·43 million), respectively.”

May combine two claims.

Unintentional injuries accounted for the most injury deaths in 2016, with 1.80 million deaths, mainly from falls, drowning, and exposure to mechanical forces.

“Unintentional injuries accounted for the most injury deaths in 2016, with 1·80 million deaths (95% UI 1·59 million to 1·89 million) composed mainly of deaths from falls (678 000 deaths [559 000–719 000]), drowning (302 000 deaths [273 000–322 000]), and exposure to mechanical forces (155 000 deaths [124 000–165 000]).”

May combine two claims.

Chronic respiratory diseases contributed 8.96% of NCD deaths in 2016, with COPD causing the most deaths from these conditions (2.93 million).

“Chronic respiratory diseases contributed 8·96% of NCD deaths in 2016, with chronic obstructive pulmonary disease (COPD) leading to the most deaths from these conditions (2·93 million deaths [95% UI 2·82 million to 3·12 million]).”

May combine two claims.

In 2016 there were 5.00 million under-5 deaths, a 33.0% decrease from 7.46 million in 2006.

“In 2016 there were 5·00 million deaths (95% UI 4·78 million to 5·23 million) in children under 5, a decrease of 33·0% (29·7–36·2) from 2006, when 7·46 million (7·27 million to 7·66 million) children died.”

May combine two claims.

Diabetes absolute deaths increased 31.1% and total YLLs increased 25.3% between 2006 and 2016, while age-standardised YLL rates decreased 2.12%.

“The absolute number of deaths and the total YLLs from diabetes both increased between 2006 and 2016 by 31·1% (95% UI 28·9–33·4) and 25·3% (23·2–27·7), respectively, while age-standardised YLL rates decreased 2·12% (0·29–3·81) over the same time period.”

May combine two claims.

Age-standardised rates did not change significantly globally for amphetamine use disorders, opioid use disorders, or other drug use disorders from 2006 to 2016.

“while rates did not change significantly on a global scale for amphetamine use disorders, opioid use disorders, or other drug use disorders over that time period.”

May combine two claims.

Age-standardised death rates fell across most neoplasms, most notably for stomach cancer (22.5% decrease to 12.6 per 100 000) and Hodgkin's lymphoma (22.4% decrease to 0.4 per 100 000).

“Age-standardised death rates fell across most neoplasms, most notably for stomach cancer (decrease of 22·5% [95% UI 20·7–24·5], to 12·6 deaths [12·3–12·9] per 100 000 in 2016) and Hodgkin’s lymphoma (decrease of 22·4% [19·9–25·3], to 0·4 deaths [0·4–0·5] per 100 000 in 2016).”

May combine two claims.

In 2016, the largest numbers of NCD deaths at Level 2 were from cardiovascular diseases (17.6 million), neoplasms (8.93 million), and chronic respiratory diseases (3.54 million).

“For NCDs in 2016, the largest number of deaths at Level 2 were caused by cardiovascular diseases (17·6 million deaths [95% UI 17·3 million to 18·1 million]) followed by neoplasms (8·93 million deaths [8·75 million to 9·09 million]), and chronic respiratory diseases (3·54 million deaths [3·40 million to 3·74 million]; table 2).”

May combine two claims.

Protein-energy malnutrition resulted in 1.22 million deaths globally from 1980 to 2016, largely in eastern sub-Saharan Africa and North Africa and the Middle East during the 1980s and in East Asia from 1995 through 2002.

“Protein-energy malnutrition resulted in 1·22 million (95% UI 670 000 to 1·81 million) deaths globally from 1980 to 2016 (figure 4C), a total that is largely composed of mortality in eastern sub-Saharan Africa and North Africa and the Middle East during the 1980s, and in countries in ast Asia from 1995 through 2002.”

May combine two claims.

In 2016, transport injuries caused 1.44 million deaths and self-harm and interpersonal violence caused 1.21 million deaths.

“In terms of number of deaths in 2016, unintentional injuries were followed by transport injuries (1·44 million deaths [95% UI 1·40 million to 1·49 million]), and self-harm and interpersonal violence (1·21 million deaths [1·11 million to 1·29 million]).”

May combine two claims.

Neonatal deaths decreased 28.9% from 3.04 million in 2006 to 2.16 million in 2016, with decreases across all mortality sources but non-significant differences for neonatal sepsis and other neonatal infections.

“Deaths in this group decreased by 28·9% (95% UI 25·5–32·1) from 2006 to 2016, from 3·04 million (2·97 million to 3·12 million) to 2·16 million (2·06 million to 2·27 million); these decreases occurred across all mortality sources, although differences were not significant for neonatal sepsis and other neonatal infections.”

May combine two claims.

Among the leading ten causes of YLLs, the median rate of change was a decrease of 2.89%, higher than the median decrease of 1.59% for causes below the leading ten.

“Among the leading ten causes of YLLs, the median rate of change was a decrease of 2·89%, higher than the median rate of change (decrease of 1·59%) for causes below the leading ten for YLLs.”

May combine two claims.

Physical violence by firearm was the largest portion (41.2%) of overall interpersonal violence in 2016, and its age-standardised death and YLL rates decreased from 2006 to 2016.

“Deaths from physical violence by firearm were the largest portion (41·2%) of overall interpersonal violence in 2016; globally, age-standardised rates for deaths and for YLLs of physical violence by firearm decreased from 2006 to 2016 by 5·34% (95% UI 0·80–9·52) and 5·19% (0·51–9·71), respectively.”

May combine two claims.

Global Zika virus disease deaths were two in 2015 and 19 in 2016.

“The global number of deaths from Zika virus disease—newly estimated for GBD 2016— was two deaths (95% UI 1–5) in 2015 and 19 deaths (4–57) in 2016.”

May combine two claims.

Natural disasters were large contributors to fatal discontinuities between 1980 and 2016, with high mortality years in South Asia, Southeast Asia, East Asia, and the Caribbean.

“Generally reflecting single large events within regions, high mortality levels from exposure to forces of nature were estimated for the years 1991 (143 000 [95% UI 65 000–222 000] deaths in South Asia), 2004 (214 000 [127 000–297 000] deaths in Southeast Asia), 2008 (141 000 [73 400–209 000] deaths in Southeast Asia, and 96 300 [15 300–185 000] deaths in East Asia), and 2010 (231 000 [36 700–443 0”

May combine two claims.

All maternal and neonatal causes of death decreased globally in both total deaths and age-standardised death rates between 2006 and 2016.

“All maternal and neonatal causes of death decreased globally in terms of both total deaths and age-standardised death rates between 2006 and 2016.”

May combine two claims.

Epidemic-prone infectious diseases caused fatal discontinuities from 1980 to 2016, with peak years in 1991, 1996, and 2014 dominated by outbreaks in western sub-Saharan Africa.

“Epidemic-prone infectious diseases also resulted in fatal discontinuities from 1980 to 2016 (figure 4B) with peak years in 1991, 1996, and 2014 dominated by mortality from infectious disease outbreaks in western sub-Saharan Africa.”

May combine two claims.

The median annualised rate of change for the leading causes of YLLs was a decrease of 2.89%, compared with 1.59% for the remainder of causes.

“The median annualised rate of change for this set of leading causes of YLLs was a decrease of 2·89%, compared with 1·59% for the remainder of causes in the hierarchy.”

May combine two claims.

In low-SDI locations, shifts in cause rankings by YLLs from 2006 to 2016 occurred beyond the top ten causes, with many CMNN causes decreasing in rank and being surpassed by NCD causes and injuries.

“In low-SDI locations, shifts in cause rankings by YLLs from 2006 to 2016 occurred beyond the top ten causes, with many CMNN causes decreasing in rank, surpassed by NCD causes and injuries.”

May combine two claims.

Absolute numbers of YLLs increased for NCDs in all but the high-SDI quintile, alongside steadily increasing life expectancy in most locations.

“While declines in age-standardised YLL rates are an indicator of progress, the absolute numbers of YLLs increased for NCDs in all but the high-SDI quintile; these increases in absolute YLLs from NCDs occurred in tandem with steadily increasing life expectancy in most locations worldwide.21”

• cites [21] 2017, Global, regional, and national under-5 mortality, adult mortality, age-specific mortality, and life expectancy, 1970–2016: a systematic analysis for the Global Burden of Disease Study 2016, Lancet sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Ischaemic heart disease was common among Latin American and Caribbean countries, ranking in the leading three causes in all countries in the region, and lower respiratory infections were commonly within the leading five causes in South Asia and sub-Saharan Africa.

“Nonetheless, ischaemic heart disease was common among countries of Latin America and the Caribbean, ranking in the leading three causes in all countries in the region, and lower respiratory infections were also commonly within the leading five causes of YLLs in South Asia and sub-Saharan Africa.”

May combine two claims.

NCDs surpassed CMNN diseases in global all-ages YLL rate in 2009 and in age-standardised rates in 1992.

“NCDs surpassed CMNN diseases in terms of global all-ages YLL rate in 2009 and earlier—in 1992—for age-standardised rates.”

May combine two claims.

Observed YLL levels much higher than expected on the basis of SDI were most commonly noted for HIV/AIDS and malaria in sub-Saharan Africa, diabetes in Oceania, and cardiomyopathy and myocarditis in eastern and central Europe.

“The opposite, where observed levels were much higher than expected on the basis of SDI, was most commonly noted in the case of HIV/AIDS and malaria in sub-Saharan Africa; diabetes mellitus, especially in Oceania; and cardiomyopathy and myocarditis, particularly in eastern and central Europe.”

May combine two claims.

Among the three broad Level 1 cause groups, rates of change for age-standardised YLLs were more heterogeneous for CMNN causes than for non-communicable causes, with the least variation for injuries.

“Between the three broad Level 1 cause groups, there was greater heterogeneity in rates of change for age-standardised YLLs for CMNN causes compared with non-communicable causes, while the least variation occurred for injuries.”

May combine two claims.

For men, conflict was a leading cause of YLLs in Syria, Yemen, and Afghanistan, and self-harm was a leading cause in Greenland and South Korea.

“A few leading causes stand out for their departure from global or regional patterns; for men these included conflict in Syria, Yemen, and Afghanistan, and self-harm in Greenland and South Korea.”

May combine two claims.

Ischaemic heart disease was the leading cause of YLLs for men in 113 countries and for women in 97 countries.

“Ischaemic heart disease was the leading cause of YLLs for men in 113 countries and for 97 countries for women, spanning both high-SDI and high-middle SDI locations as well as many lower SDI locations, such as Kyrgyzstan.”

May combine two claims.

In China, lung, liver, stomach, and oesophageal cancers were within the leading ten causes of YLLs and each caused higher than expected YLLs on the basis of SDI.

“In China, a number of cancers—lung cancer, liver cancer, stomach cancer, and oesophageal cancer—were within the leading ten causes of YLLs, and each of these caused higher than expected levels of YLLs on the basis of SDI; from 1·70 times higher for lung cancer to 5·43 times higher for liver cancer.”

May combine two claims.

Other leading causes with observed YLLs notably lower than expected included neonatal preterm birth complications in south and southeast Asia and cerebrovascular disease in western Europe.

“Other leading causes for which observed YLLs were notably lower than expected included neonatal preterm birth complications in many countries of both south Asia and southeast Asia, and cerebrovascular disease in western Europe.”

May combine two claims.

There was no change from 2006 to 2016 in the leading three causes of YLLs for high-SDI, high-middle-SDI, and middle-SDI locations, although age-standardised YLL rates decreased for each of these causes.

“There was no change from 2006 to 2016 in the leading three causes of YLLs for high-SDI, high-middle-SDI, and middle-SDI locations, although age-standardised YLL rates decreased for each of these causes (figure 7A–C).”

May combine two claims.

NCD age-standardised YLL rates declined for each SDI quintile from 1980 to 2016, with the slowest decline at low SDI and the fastest at high SDI.

“NCD age-standardised YLL rates also declined for each SDI quintile from 1980 to 2016, but rates of decline were slowest at low SDI and fastest for high SDI.”

May combine two claims.

Over the 37-year period, global HIV/AIDS deaths rose from 4224 in 1980, peaked in 2005 at 1.91 million, and declined to 1.03 million in 2016.

“Over the 37-year period analysed here, global deaths due to the HIV/AIDS epidemic rose from 4224 deaths (95% UI 2842–6274) in 1980, peaked in 2005 at 1·91 million (1·81 million to 2·01 million), and declined to 1·03 million (0·99 million to 1·08 million) in 2016.”

May combine two claims.

Across much of sub-Saharan Africa, the leading cause of YLLs for both sexes varied between HIV/AIDS, malaria, diarrhoeal diseases, and lower respiratory infections; for men in the Central African Republic, tuberculosis was the leading cause.

“Across much of sub-Saharan Africa, the leading cause for both men and women varied between HIV/AIDS, malaria, diarrhoeal diseases, and lower respiratory infections; however, for men in the Central African Republic, tuberculosis was the leading cause.”

May combine two claims.

The causes contributing the most YLLs varied with location, although regional and SDI-associated patterns were evident.

“however, the causes that contributed the most YLLs varied with location—although regional patterns and patterns associated with SDI were also evident.”

May combine two claims.

The largest reduction in age-standardised YLL rates for high-SDI locations came from road injuries, which fell 25.2%, while the largest increase was for drug use disorders, which rose 18.9%.

“The largest reduction in age-standardised YLL rates for high-SDI locations came from road injuries, which fell by 25·2% (95% UI 22·5–27·6), while the largest increase was for drug use disorders (rising by 18·9% [13·8–23·3]).”

May combine two claims.

There was considerably more heterogeneity in trends by SDI quintile and GBD region or location, with 36 countries having significant increases in age-standardised YLLs across 3 or more Level 3 causes.

“For SDI quintiles, and by GBD regions or locations, there was considerably more heterogeneity in trends—eg, 36 countries had significant increases in age-standardised YLLs across 3 or more of Level 3 causes.”

May combine two claims.

In 2016, the leading 30 causes of all-age YLLs for high-SDI locations were predominantly NCDs and injuries, with one CMNN cause, lower respiratory infections, within the leading 30.

“In 2016, the leading 30 causes of all-age YLLs for high-SDI locations were predominantly from NCDs and injuries, with one CMNN cause—lower respiratory infections—within the leading 30 causes (figure 7A).”

May combine two claims.

Age-standardised YLL rates due to injuries varied least across quintiles and over time but were highest in the low-SDI quintile.

“Age-standardised YLL rates due to injuries varied the least across quintiles and over time but were highest in the low-SDI quintile.”

May combine two claims.

The transition was characterised by declining CMNN disease rates in all SDI quintiles, with faster declines in lower-SDI quintiles.

“This transition was characterised by declining rates of CMNN diseases in all SDI quintiles, with faster rates of decline in lower-SDI quintiles.”

May combine two claims.

Injury age-standardised YLL rates declined in each SDI quintile, although this was punctuated by large increases in some years due to fatal discontinuities.

“Injury age-standardised YLL rates declined in each SDI quintile, although this was punctuated by large increases in some years due to fatal discontinuities from conflict and terrorism, disasters, and epidemics.”

May combine two claims.

In all SDI quintiles, age-standardised NCD rates decreased, with the slowest decline in low SDI and the fastest in high SDI.

“In all quintiles, age-standardised rates for NCDs have decreased; by contrast with trends for CMNN causes, the pace of decline for NCD rates was slowest in low SDI and fastest in high SDI.”

May combine two claims.

Global reductions in age-standardised death rates due to hepatitis C have largely stagnated since 2010 despite the availability of a highly effective cure, and in the USA total hepatitis C deaths have surpassed deaths from all other notifiable infectious diseases.

“Globally, reductions in age-standardised death rates due to hepatitis C have largely stagnated since 2010 even though a highly effective cure is available—and in the USA, total deaths from hepatitis C have now surpassed deaths from all other notifiable infectious diseases,49”

• cites [49] McMichael, 2006, Climate change and human health: present and future risks, Lancet sourceNot checked yet

  Not checked in the free run.

May combine two claims.

The new SRS data changed India results and, by changing regression coefficients, also modified estimates for other locations with lower data-quality star ratings.

“In some cases, the new SRS data over a 10-year period not only changed results for India but, through changing coefficients in regression models, also modified estimates for other locations that were lower in the stars system of data quality rankings developed for GBD 2016.”

May combine two claims.

Global Zika virus deaths were estimated at 2 (95% UI 1-5) in 2015 and 19 (4-57) in 2016.

“The global number of Zika virus deaths was comparatively small—estimated at two (95% UI 1–5) in 2015 and 19 (4–57) in 2016.”

May combine two claims.

The GBD drug-resistant tuberculosis share is lower than WHO's 13.9% for the same year because of different composition of drug-resistant types.

“Due to the different composition of drug-resistant types, these numbers are lower than the 13·9% of drug-resistant tuberculosis deaths (multidrug-resistant and rifampicin-resistant tuberculosis combined) among all tuberculosis deaths reported by WHO for the same year.6”

• cites [6] 2016Not checked yet

  Not checked in the free run.

May combine two claims.

Malaria changes in India were most pronounced and were driven by inclusion of improved SRS mortality data, which lowered estimates in all years, especially earlier years.

“Changes in India were most pronounced and driven by inclusion of the much-improved SRS mortality data, which lowered estimates in all years, but particularly earlier years.”

May combine two claims.

Updates to malaria data and methods in GBD 2016 changed estimates of both recent malaria mortality and trends over time.

“Updates in the data collated and improvements to the methodological approach for malaria mortality estimation in the GBD 2016 iteration have resulted in changes in estimates of both contemporary mortality and trends through time.”

May combine two claims.

Alzheimer's disease and other dementias and chronic kidney disease ranked fourth and 11th among leading causes of death globally and are increasing sources of health burden, particularly in low-SDI to middle-SDI locations.

“Two causes—Alzheimer’s disease and other dementias, and chronic kidney disease, ranked fourth and 11th, respectively, among the leading causes of death globally—are increasing sources of health burden, particularly for low-SDI to middle-SDI locations; however, the SDG agenda offers at best a minimal platform for drawing attention to the health care and monitoring needs of these conditions.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Data quality improved in some large countries, including Brazil, China, and India.

“The improvements in data quality in some large countries, such as Brazil, China, and India, is particularly encouraging.”

May combine two claims.

Global deaths from other pharynx cancers in 2010 increased by 72.3% for men and 73.8% for women compared with GBD 2015, mostly from India.

“For example, estimates of deaths from other pharynx cancers in 2010 at the global level increased by 72·3% for men and by 73·8% for women; most of this increase comes from India.”

May combine two claims.

The proportional decline in malaria deaths between 2006 and 2015 is estimated at 27.9% (95% UI 8.71-42.8), smaller than WHO's estimated 42% (33-55).

“As such, the proportional decline in malaria deaths between 2006 and 2015 is now estimated to be 27·9% (95% UI 8·71–42·8), which is smaller than the WHO estimated decline of 42% (33–55).7”

• cites [7] 2016Not checked yet

  Not checked in the free run.

May combine two claims.

Global deaths from Hodgkin's lymphoma in 2010 increased by 28.1% for men and 16.8% for women compared with GBD 2015, mainly from sub-Saharan Africa due to more registry data.

“Estimates for deaths from Hodgkin’s lymphoma in 2010 increased by 28·1% for men and by 16·8% for women compared with GBD 2015 at the global level. This difference comes mainly from an increase in the mortality estimates for sub-Saharan Africa, where the inclusion of more registry data for Nigeria, Uganda, and other locations changed the estimates.”

May combine two claims.

Global malaria estimates for the most recent years are slightly higher than in GBD 2015, while earlier years, including the early-2000s peak, are lower.

“At the global level, estimates for malaria in the most recent years are slightly higher compared with those in GBD 2015, but earlier years were estimated to be lower, including the years of peak global malaria mortality in the early 2000s.”

May combine two claims.

The revised tuberculosis modelling strategy did not majorly change general conclusions on global tuberculosis epidemiology, although global deaths were slightly higher than in GBD 2015 for all years.

“This, together with the addition of SRS data for India and changes to the mortality envelope, has not resulted in major changes to our general conclusions on the global epidemiology of tuberculosis, although the number of deaths at the global level was slightly higher than that of GBD 2015 for all years.”

May combine two claims.

Global health patterns are changing, with more rapid declines in communicable, maternal, neonatal, and nutritional conditions than in other diseases and injuries.

“Patterns of global health are clearly changing, with more rapid declines in CMNN conditions than for other diseases and injuries.”

May combine two claims.

Egypt, Thailand, and Turkey each had more than 50% of deaths assigned to major garbage categories.

“By contrast, some systems, such as those in Egypt, Thailand, and Turkey, have more than 50% of deaths assigned to major garbage categories.”

May combine two claims.

Estimated tuberculosis deaths increased significantly in several African countries; Nigeria and Zambia had more than twice the estimated deaths in 2015.

“The estimated number of deaths in several African countries have significantly increased; Nigeria and Zambia were notable with more than twice the number of estimated deaths in 2015.”

May combine two claims.

Global deaths from other neoplasms in 2010 increased by 20.3% for men and 5.1% for women compared with GBD 2015, mainly due to redistribution of myelodysplastic syndrome.

“Compared with in GBD 2015, estimated deaths in 2010 due to other neoplasms increased by 20·3% for men and by 5·1% for women. This increase was mainly due to a change in redistribution of myelodysplastic syndrome, which was redistributed to the leukaemia subcauses for GBD 2015, and to other neoplasms for GBD 2016.”

May combine two claims.

In Western Europe, overall early neonatal death rates from neonatal disorders are similar across the region, but death rates within Level 3 causes differ by as much as three-fold.

“Western Europe is an example where the overall death rate due to neonatal disorders in the early neonatal period is quite similar across the region, but there is as much as a three-fold difference in death rates within Level 3 causes (eg, neonatal preterm birth complications, cardiovascular diseases).65,66”

• cites [65] Delnord, 2017, Variations in very preterm birth rates in 30 high-income countries: are valid international comparisons possible using routine data?, BJOG sourceNot checked yet

  Not checked in the free run.

• cites [66] Roth, 2017, Global, Regional, and National Burden of Cardiovascular Diseases for 10 Causes, 1990 to 2015, J Am Coll Cardiol sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

The MCEE group reports estimates for 15 cause groups of child death for 194 countries for 2000-15.

“The latest assessment from the Maternal and Child Epidemiology Estimation (MCEE) group reports estimates for 15 cause groups of child death for 194 countries for the period 2000–15,18”

• cites [18]Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

All authors had full access to the data and final responsibility for the decision to submit for publication.

“All authors had full access to the data in the study and had final responsibility for the decision to submit for publication.”

Probably not a substantial claim: may not pass S3 (load-bearing).

WHO's Global Health Estimates programme published estimates for 176 causes of death for 183 countries from 2000 to 2015.

“while the Global Health Estimates (GHE) programme through WHO recently published estimates for 176 causes of death for 183 countries from 2000 to 2015.19”

• cites [19] 2016Not checked yet

  Not checked in the free run.

May combine two claims.

Independent CODEm models were run for each cause by sex and separately for countries with and without extensive complete vital registration data.

“Independent CODEm models were run for each cause of death by sex, and separately for countries with and without extensive complete VR data.”

May combine two claims.

CODEm models were combined into a weighted ensemble, with models performing best on out-of-sample prediction error for levels and trends weighted highest.

“These models were then combined into a weighted ensemble model, with models performing best on out-of-sample prediction error of both levels and trends weighted highest.”

Classified as probably not a substantial claim.

The entire GBD time series was re-estimated to ensure comparisons used a consistent dataset and methods.

“As with each annual cycle of GBD, the entire time series was re-estimated to ensure that all comparisons are made using a consistent dataset and methods; these results, therefore, supersede all previously published GBD cause of death estimates.”

May combine two claims.

Over the past 37 years, rates of communicable, maternal, neonatal, and nutritional diseases declined across all SDI quintiles, with faster than expected gains for many locations relative to their SDI.

“The past 37 years have featured declining rates of communicable, maternal, neonatal, and nutritional diseases across all quintiles of SDI, with faster than expected gains for many locations relative to their SDI.”

• AssumptionRests on an assumption: The GBD cause hierarchy's categories at each level are mutually exclusive and collectively exhaustive.
• AssumptionRests on an assumption: Each death can be assigned to a single underlying cause according to ICD principles.

Probably not a substantial claim: may not pass S3 (load-bearing).

The funder had no role in study design, data collection, data analysis, data interpretation, or writing the report.

“The funder of the study had no role in the study design, data collection, data analysis, data interpretation, or the writing of the report.”

Probably not a substantial claim: may not pass S3 (load-bearing).

GBD 2016 followed the GATHER guidelines for development and documentation.

“As in GBD 2015, we follow the Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER) for the development and documentation of GBD 2016 (appendix 1 p 292).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Rapid progress to reduce mortality is possible for some causes, as shown by documented declines in cardiovascular disease death rates in central Europe and decreasing malaria mortality in eastern sub-Saharan Africa.

“Rapid progress to reduce mortality is possible for some causes, as evidenced by previously documented declines in central Europe for cardiovascular disease death rates or decreasing mortality from malaria in eastern sub-Saharan Africa.3”

• cites [3] Wang, 2016, Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980–2015: a systematic analysis for the Global Burden of Disease Study 2015, Lancet sourceNot checked yet

  Not checked in the free run.

May combine two claims.

International declarations and agreements for development and health have generally focused on CMNN-related indicators, and this focus is reflected in the rapid reductions in CMNN diseases achieved by lower-SDI locations.

“International declarations and agreements for development and health have attracted political actions at the highest level but have also generally focused on indicators related to CMNN causes. This focus is reflected in the rapid reductions in CMNN diseases achieved by lower-SDI locations over the past four decades.”

May combine two claims.

Global progress in mortality has been neither universal nor uniform.

“Globally, progress has been neither universal nor uniform.”

May combine two claims.

The quality of data available for estimating global causes of death improved, and with the new rating system 50 countries had a higher star rating in 2010-16 than their overall rating since 1980.

“The quality of data available for estimating the global causes of death has improved; with the new rating system reported here, 50 countries had a higher star rating in the most recent period (2010–16) compared with their overall rating since 1980.”

May combine two claims.

Increasing YLLs and higher chronic disease prevalence will drive rising health service needs and pose financial, workforce, and managerial challenges for health systems.

“Increasing numbers of YLLs and the associated higher prevalence of chronic diseases will drive rising needs for health service provision and pose substantial financial, workforce capacity, and managerial challenges for health systems.”

May combine two claims.

Until the 2015 adoption of the SDGs, NCDs were not strong priorities in international declarations and did not receive equivalent political commitment in many locations.

“However, until the 2015 adoption of the SDGs, NCDs were not strong priorities in these declarations and did not receive equivalent levels of political commitment in many locations.45–47”

• cites [45] Bollyky, 2015, Understanding the relationships between noncommunicable diseases, unhealthy lifestyles, and country wealth, Health Aff (Millwood) sourceNot checked yet

  Not checked in the free run.

• cites [46] Anderson, 2016, Tobacco control progress in low and middle income countries in comparison to high income countries, Int J Environ Res Public Health sourceNot checked yet

  Not checked in the free run.

• cites [47] Ly, 2016, Rising mortality associated with hepatitis C virus in the United States, 2003–2013, Clin Infect Dis sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Past studies show that access to high-quality personal health care affects both communicable and non-communicable diseases.

“Past studies show the effect of access to high-quality personal health care on both communicable and non-communicable diseases, highlighting the importance of prioritising personal health-care access and quality for all populations across the development spectrum.39–42”

• cites [39] NolteNot checked yet

  Not checked in the free run.

• cites [40] Barber, 2017, Healthcare Access and Quality Index based on mortality from causes amenable to personal health care in 195 countries and territories, 1990–2015: a novel analysis from the Global Burden of Disease Study 2015, Lancet sourceNot checked yet

  Not checked in the free run.

• cites [41] Ford, 2007, Explaining the decrease in U.S. deaths from coronary disease, 1980–2000, N Engl J Med sourceNot checked yet

  Not checked in the free run.

• cites [42] Tunstall-Pedoe, 1999, Contribution of trends in survival and coronary-event rates to changes in coronary heart disease mortality: 10-year results from 37 WHO MONICA Project populations, Lancet sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Large differences in neonatal cause-of-death patterns between countries with nearly the same SDI in the same GBD region are likely not real and are suspected to reflect variation in medical culture in assigning neonatal deaths.

“There is no reason to believe that these large differences in the causes of deaths in the first month of life between countries with nearly the same SDI in the same GBD region are real; rather, we strongly suspect there is variation in medical culture in how neonatal deaths are assigned.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Dengue and its four serotypes have continued to increase and expand globally.

“The continued global increase and expansion of dengue and its four serotypes is a potential indicator of the complex changes that are underway and might be partly related to changes in climate in addition to other factors.61”

• cites [61] Messina, 2014, Global spread of dengue virus types: mapping the 70 year history, Trends Microbiol sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

It is unclear whether bias from underestimated IARC MIRs primarily affects IARC incidence estimates or site-specific mortality estimates.

“Due to the complexities of the IARC estimates, if the bias in their estimates from underestimated MIRs primarily affects their estimates of incidence or the site-specific mortality estimates is unclear.”

Probably not a substantial claim: may not pass S3 (load-bearing).

IARC Globocan estimates do not currently meet GATHER guidelines.

“The Globocan project at IARC produces estimates of major cancer types on a periodic basis for 184 countries; IARC estimates do not currently meet GATHER guidelines.15”

• cites [15] 2017Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

As recognized in the GHE 2015 technical paper, GBD remains the only source of comprehensive uncertainty estimates for mortality by cause.

“The GBD study, as recognised in the GHE 2015 technical paper, remains the “only source of comprehensive uncertainty estimates for mortality by cause.”19”

• cites [19] 2016Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Improvements to broader social determinants of health might be less developed in low-SDI locations; for example, low-income and middle-income countries lag in implementing evidence-based tobacco control regulation.

“In addition to the role of health care, improvements to the broader social determinants of health might be less developed in low-SDI locations—eg, low-income and middle-income countries lag behind in implementing evidence-based tobacco control regulation.48”

• cites [48] Patz, 2005, Impact of regional climate change on human health, Nature sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

A low level of identified garbage coding for a location does not necessarily indicate quality or accuracy in cause-of-death certification.

“Seventh, a low level of identified garbage coding for a given location does not necessarily indicate quality or accuracy in cause of death certification.”

May combine two claims.

GBD 2015 mortality-to-incidence ratios did not capture likely effects of worse treatment access in lower-SDI settings, and the revised MIR approach better captures the relationship between MIR and health-system access and quality of care.

“GBD 2015 estimated MIRs did not capture the likely effects of worse access to treatment in lower-SDI settings; we have revised the MIR data-inclusion and modelling approach to better capture the relationship observed in high-quality registry data between MIR (and implicitly 5-year survival) and health-system access and quality of care.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Additional sources of uncertainty may not be captured, such as uncertainty for model covariates other than the HIV crude death rate.

“Tenth, additional sources of uncertainty might not be captured, such as for the covariates used in the models with the exception of the HIV crude death rate.”

Probably not a substantial claim: may not pass S3 (load-bearing).

WHO's Global Health Estimates for cause of death cover 2000-2015 for 183 countries and 176 causes.

“WHO has produced cause of death Global Health Estimates (GHE) at the country level with the most recent spanning from 2000 to 2015 for 183 countries and 176 causes of death.19”

• cites [19] 2016Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

The SDGs focus on tracking hepatitis B incidence rather than hepatitis C mortality.

“yet the SDGs are focused on tracking hepatitis B incidence (SDG 3.3.4).”

• cites SDG 3.3.4Not checked yet

  Not checked in the free run.

May combine two claims.

Verbal autopsy data sources vary substantially in training and instrument, which may reduce comparability of cause-of-death data between locations.

“Fifth, sources of VA data vary substantially in terms of the training provided and the instrument used in collecting the data, which might reduce the comparability of cause of death data between locations.”

May combine two claims.

GBD results combine data and estimation; recent-year estimates rely more on modelling, as shown by larger median uncertainty intervals from 2012 to 2016, and the same is true for locations with low data completeness.

“Finally, GBD results are necessarily a combination of data and estimation. Due to lags in reporting, estimates for the most recent years rely more on the modelling process—evidenced by larger median UI by year between 2012 and 2016— as do estimates for locations with low levels of data completeness.”

May combine two claims.

Garbage code redistribution is vital to GBD 2016 results; the methods could theoretically contribute bias, but no evidence of such bias was identified.

“Sixth, our approach to garbage code redistribution is vital to the results presented in GBD 2016—although our methods of redistribution could theoretically contribute bias, we have identified no evidence of this.3”

• cites [3] Wang, 2016, Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980–2015: a systematic analysis for the Global Burden of Disease Study 2015, Lancet sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Both vital registration and verbal autopsy data depend on accurate assignment of underlying cause of death, which is complicated by multimorbidities.

“Second, both VR data and VA data sources depend on how accurately underlying cause of death is assigned and this is complicated by multimorbidities.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The rapid declines in CMNN conditions reflect the global health community's intense focus over recent decades on improving child survival and reducing pregnancy-related risks.

“This is a laudable and welcome reflection of the intense focus of the global health community over the past several decades on improving child survival and reducing pregnancy-related risks.”

May combine two claims.

In rigorous validation tests, verbal autopsy performs well for some causes and poorly for others.

“Nevertheless, these data are still based on VA, which in rigorous validation tests performs well for some causes and not others.63”

• cites [63] Murray, 2011, Population Health Metrics Research Consortium gold standard verbal autopsy validation study: design, implementation, and development of analysis datasets, Popul Health Metr sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

No standardised empirical approach has been developed for selecting causes to use negative binomial modelling.

“Eighth, while some causes use negative binomial modelling approaches to improve estimation with over-dispersed data, we have not yet developed a standardised empirical approach for selecting causes to use this method.”