Registry & Real-World Data Digest — Issue 01

Lei Yang, Ning Kang, Tao Xue, et al.

Summary: Researchers used data from 172,845 lung cancer patients in Beijing, diagnosed between 2003 and 2022, to study how long-term exposure to two types of air pollution — fine particles (PM2.5) and ground-level ozone — related to the risk of dying. Both types of pollution were independently linked to higher death rates. The study also used a modeling approach to estimate what might have happened if pollution levels had stayed at 2013 levels. It estimated that PM2.5 reductions since 2013 were linked to roughly 13,229 fewer deaths and an average gain of about 7.8 months of expected survival over five years — but that rising ozone levels canceled out about 40% of that estimated benefit.

Why it matters: China's major push to reduce air pollution has brought down PM2.5 levels, and this study tries to put numbers on what that may have meant for people already living with lung cancer — a group particularly vulnerable to the effects of air pollution. The finding that ozone, which can increase when certain other pollutants are reduced, has partially erased those gains sends a practical signal to policymakers: reducing one type of pollution while another rises may not produce the full health benefit expected.

What to keep in mind: This study observed patterns in registry data; it did not randomly assign people to different pollution levels, so it cannot prove that pollution directly caused the changes in survival. The specific numbers — 13,229 avoided deaths, 7.8 months of gained survival — are outputs of a mathematical model built on several assumptions, and should be understood as estimates rather than measured facts. Many other things changed between 2013 and 2022 in Beijing, including cancer treatments, screening practices, and broader social conditions, and it is difficult to fully separate the effect of cleaner air from all of those changes. The Beijing Cancer Registry is one of China's better-resourced registries, but important clinical details — such as cancer stage, treatment received, and smoking history — are not fully captured, which limits how confident we can be in the precise figures.

Relevance: Registry and Real-World Data Methods

Source: https://pubmed.ncbi.nlm.nih.gov/42778299/

---

Xu Niu, Tangxing Jiang, Wen Zheng, et al.

Summary: This study compared two medications — lidocaine and amiodarone — used during a specific type of life-threatening cardiac arrest (where the heart's electrical rhythm is dangerously abnormal and does not respond to electric shocks). Researchers drew on registry data from 40 hospitals across China, identifying 612 patients out of more than 32,000 cardiac arrest cases who met the study criteria. They found no statistically significant difference between the two drugs in survival to hospital discharge, return of a stable heartbeat, or favorable brain function outcomes.

Why it matters: Medical guidelines currently allow doctors to choose either drug in this situation, largely based on studies of cardiac arrests that happened outside a hospital. Cardiac arrests inside hospitals are different in important ways — patients are already being monitored, medical staff respond faster, and the underlying causes differ. This study provides real-world evidence specifically for in-hospital cardiac arrest, offering some reassurance that the guideline's flexibility on drug choice is not obviously wrong in this setting.

What to keep in mind: The most important limitation is that only 612 patients out of more than 32,000 qualified for this analysis — a small group for this kind of comparison. The results are consistent with either drug being better, worse, or truly equivalent; the study simply was not large enough to detect a meaningful difference if one exists. Because patients were not randomly assigned to a drug, the choice of medication may have been influenced by factors the researchers could not measure, such as how the patient appeared to be doing at the time of the arrest or which drugs were available at that hospital. The study was also conducted partly during the COVID-19 pandemic, which may have affected how cardiac arrests were managed. The findings apply most directly to Chinese hospital settings, where lidocaine has historically been used more often than in Western countries.

Relevance: Registry and Real-World Data Methods

Source: https://pubmed.ncbi.nlm.nih.gov/42767592/

---

James White, Sarah J Hampton, Shaun Greene

Summary: Researchers analyzed 46 deaths recorded in a Victorian (Australia) clinical registry — the Emerging Drugs Network of Australia-Victoria, known as EDNAV — between September 2020 and December 2025, out of more than 5,000 total presentations. Almost all of the people who died were male (89%), with a median age of 33. Nearly every case (91%) involved more than one substance, with a median of three substances detected. Methamphetamine and heroin were found most often. Two-thirds of deaths involved cardiac arrest before reaching hospital, and most people had already suffered irreversible brain injury by the time they arrived at emergency care. Newer synthetic drugs appeared in fewer than 8% of detections and were never the only substance found.

Why it matters: The near-universal pattern of mixing substances — particularly combining stimulants like methamphetamine with sedatives like heroin or benzodiazepines — makes harm reduction more complicated than approaches targeting a single drug. Because most deaths had already occurred or were effectively irreversible before hospital arrival, the study points toward community-based responses as the most realistic way to reduce deaths: wider distribution of naloxone (a medication that can reverse opioid overdoses), bystander training in resuscitation, and faster emergency response. The registry also shows that advanced laboratory testing can detect drug combinations that standard hospital screening misses.

What to keep in mind: This report covers only 46 deaths — a small number accumulated over five years — which limits how much can be reliably concluded from patterns within the group. The registry captures people who reached the emergency department; the majority of deaths in this cohort had already happened outside the hospital, meaning the registry only partially represents the full picture of drug-related fatality in Victoria. Detecting a substance in someone's system confirms they were exposed to it, but it does not prove that substance was the main cause of death; when multiple drugs are present, it is very difficult to determine which one, or which combination, was most responsible. The findings reflect one Australian state and may not apply to other regions or countries.

Relevance: Registry and Real-World Data Methods

Source: https://pubmed.ncbi.nlm.nih.gov/42716535/

---

Miriam G Wadström, Anders Stålman, Tobias Wörner, et al.

Summary: Swedish researchers used linked national registry data to study 4,633 people who had a type of minimally invasive hip surgery — hip arthroscopy — between 2006 and 2018, comparing them with more than 23,000 similar people from the general population who did not have the procedure. People who had the surgery were more likely to be employed, earn higher incomes, and have more education than average. Despite this, they also took substantially more compensated sick leave — both before and after surgery — than the comparison group. The procedure was far more common in and around major cities than in rural areas.

Why it matters: The concentration of this surgery among higher-income, better-educated people raises a question about fairness: is access to this procedure driven mainly by clinical need, or also by the ability to navigate the healthcare system? The geographic clustering around cities suggests that where someone lives may affect whether they receive this treatment — which could reflect where specialist surgeons practice, local referral habits, or unequal access. The high sick leave burden, both before and after surgery, also has real economic implications. These findings, drawn from Sweden's uniquely comprehensive linked national registries, would be very difficult to produce in countries without similar data infrastructure.

What to keep in mind: This study describes patterns; it does not explain why they exist. The association between higher income and receiving this surgery could reflect better access to specialist care, the types of physical activities more common among higher-income groups, or differences in how people from different backgrounds seek medical care — the data cannot separate these possibilities. Crucially, the registries used here contain administrative and employment records but not clinical information: there is no data on diagnosis, how severe each person's symptoms were, what treatments they tried before surgery, or how well the surgery worked. Without that clinical context, it is impossible to say whether people with lower incomes are being undertreated or whether the utilization differences are clinically appropriate. The sick leave data after surgery cannot distinguish between normal recovery time and poor surgical outcomes.

Relevance: Registry and Real-World Data Methods

Source: https://pubmed.ncbi.nlm.nih.gov/42776610/

---

Kristina Anna Djupvik Aakvik, Silje Dahl Benum, Marjaana Tikanmäki, et al.

Summary: Norwegian researchers linked physical activity measurements from 12,896 participants in the HUNT Study — a large, long-running population health study — with birth records from Norway's national birth registry. Participants wore motion-tracking devices to measure how active they actually were (rather than relying on self-report). The analysis found that being born earlier than expected — at a shorter gestational age — was modestly associated with lower physical activity levels, particularly for light activity in females and total activity in adolescents. Birth weight, on its own, showed no association with how active people were later in life.

Why it matters: Whether the circumstances of birth have lasting effects on health behaviors like physical activity is an important public health question, because activity levels influence the risk of many chronic diseases across a lifetime. The methods used here — connecting a high-quality population study to a national birth registry — offer a useful model for life-course research that other registry systems could replicate. For anyone interested in how registry data can be linked across time and data sources, this study is a practical illustration of what is possible with well-maintained national records.

What to keep in mind: The associations found were statistically detectable but very small in practical terms — the difference in activity linked to being born a week earlier or later was minimal across most of the population studied. The study combined people aged 13 to 52 into a single analysis, a very wide range that spans very different life stages, and patterns in birth and obstetric care changed substantially over the decades when these participants were born. Very premature births — the group most studied in other research — were likely underrepresented here, since many very premature babies born in earlier decades did not survive to adulthood. The analysis adjusted for age and sex but not for many other factors — such as a mother's health, family income, or social circumstances — that could both affect the timing of birth and shape how active someone becomes later in life. These unmeasured factors mean the study cannot establish that gestational age itself caused the differences in activity. The primary scientific contribution is the demonstration of how registry linkage can be done, rather than a definitive finding about birth and physical activity.

Relevance: Registry and Real-World Data Methods

Source: https://pubmed.ncbi.nlm.nih.gov/42776930/