The research, led by Xiaoying Zhu of Nanjing University of Chinese Medicine and the Nossal Institute for Global Health at the University of Melbourne, together with colleagues including Daniel Ll Strachan, Shenglan Tang, Tiara Marthias, Ajay Mahal and Barbara McPake, is built on face-to-face, semi-structured interviews with 21 physicians sampled from two hospitals with deliberately different incentive structures. Conducted in 2024, the interviews probed how doctors experience and respond to the Diagnosis-Related Group reform, known globally as DRG payment. The analysis followed an inductive thematic approach before applying a deductive framework, allowing the researchers to surface patterns in the physicians’ own words before testing them against established behavioural theory. The result is one of the most granular portraits to date of what health economists call cream-skimming: the systematic selection of profitable patients and the subtle exclusion of those whose care is expected to cost more than the hospital will be paid.

DRG payment systems, first developed in the United States and now deployed in dozens of countries, were designed to curb the runaway costs of fee-for-service medicine. Instead of paying hospitals for every test, drug and bed-day, the payer assigns each admission to a fixed-price category based on the diagnosis, so the hospital receives a predetermined sum regardless of how much the treatment actually costs. In theory, this pushes hospitals to become more efficient, eliminate unnecessary procedures and shorten stays. But the same mechanism creates a structural hazard: a patient whose actual treatment exceeds the fixed price becomes a financial loss, and a patient who needs less than the fixed price generates a surplus. In China, where the government has simultaneously pledged to advance healthcare equity while rolling out DRG reform nationwide, the tension between fiscal discipline and fair access has never been sharper.

The Nanjing interviews reveal that the phrase ‘more work means less pay’ is not a complaint muttered in passing but an operating principle that shapes admission decisions. Physicians described assessing, often within minutes, whether the expected DRG reimbursement would cover the anticipated treatment costs of a particular patient. Clinically complex cases — those with multiple comorbidities, unclear diagnoses, or likely prolonged stays — emerged as the paradigmatic financially risky admissions. The study found that patient selection was not framed by the doctors as greed or malice, but as an adaptive response: a rational adjustment to incentives that penalize exactly the kind of resource-intensive care that clinical complexity demands. When the payment rule makes complexity costly, the doctors’ calculus bends toward avoidance, whether they want it to or not.

What distinguishes this study is its effort to dissect the anatomy of the decision rather than merely cataloguing its existence. The researchers found that physicians’ choices were shaped by three interacting dimensions: capability, opportunity and motivation. Capability captured the physician’s confidence in managing financially risky cases under the constraints of the fixed payment — a doctor experienced in the relevant specialty, or familiar with the DRG category’s payment boundary, may feel able to admit a patient whose costs would overshoot the reimbursement. Opportunity encompassed the physical and social conditions that allowed a physician to balance financial risk against quality of care, including the availability of referral channels, the support of the hospital environment, and the flexibility to structure a treatment plan that keeps costs within the payment envelope. Motivation concerned the value physicians placed on different possible outcomes — the welfare of the patient, the fulfilment of professional obligations, and the recognition their decisions would earn within the hospital’s performance system.

The alignment of these three dimensions proved decisive. Physicians were most likely to admit a financially risky patient when all three conditions held simultaneously: when they believed the patient could be effectively managed under DRG payment, when they expected their decision to produce meaningful and recognised outcomes, and when they valued those outcomes, including the patient’s welfare and their own sense of professional duty. Remove any one of the three — sap a doctor’s confidence that the case can be managed without a punishing loss, block the pathways that would make careful management possible, or hollow out the recognition and reward for doing the right thing — and the willingness to accept a costly patient collapses. This finding reframes cream-skimming not as a fixed character flaw of physicians but as a system-dependent behaviour, sensitive to how hospitals structure support, performance metrics and clinical pathways.

The implications reach far beyond Nanjing. DRG reform is among the most ambitious payment experiments in modern health policy, and China’s version is being implemented at a scale no other health system has attempted, covering thousands of hospitals and hundreds of millions of insured patients. Studies from European and East Asian systems have repeatedly documented the same perverse incentives — upcoding of diagnoses to reach higher-paying categories, premature discharges, and the quiet diversion of complex patients to other institutions. But the Nanjing study adds a crucial layer of mechanism, showing how the behaviour is mediated by doctors’ perceptions of financial risk and by the institutional scaffolding that either absorbs or amplifies that risk. Whether a high-cost patient is admitted may depend less on the payment rule itself than on whether the hospital has created the conditions under which a physician can manage the case without absorbing the loss personally or as a department.

The authors’ conclusions point toward a set of concrete correctives. Improving equitable access, they argue, requires realigning hospital performance metrics with clinical need rather than purely with financial outcomes, so that physicians who take on difficult patients are recognised and rewarded rather than penalized. It requires strengthening referral and post-acute care pathways, so that complex cases can move through the system in ways that distribute cost and responsibility rather than concentrating both on a single admission. And it requires reducing reliance on individual risk-based decision-making through team-based support, spreading the burden of financially risky cases across the institution instead of leaving each physician to weigh it alone at the bedside. In each case, the goal is the same: to change the ecology in which the admission decision is made, rather than to exhort doctors to behave better within an ecology that punishes them for it.

For patients, the stakes are stark. A physician who declines or deflects a complex admission does not simply shift an accounting problem; the patient may face longer waits, fragmented care, or repeated referrals before reaching someone willing to treat them. For the most vulnerable — the elderly with multiple chronic conditions, the poor with advanced disease — the cumulative effect of thousands of individual admission decisions, each nudged by the same financial logic, is a system that quietly rationed care by complexity. That is the opposite of what health equity requires, and it is why the study’s authors frame their work as an urgent input into China’s equity commitments. The research received no specific grant from any funding agency; Zhu’s PhD program was supported by a China Scholarship Council–University of Melbourne PhD Scholarship.

The study is a reminder that health payment reform is never merely a technical exercise in accounting. Every fixed price, every performance indicator, every cost-allocation rule lands eventually on a human decision made under pressure — and when the rule says that the hardest patients are the least rewarding, the system must decide whether it intends that outcome or will redesign itself to prevent it. The Nanjing physicians interviewed for this study did not invent the dilemma; they inherited it from the architecture of the reform. What the study makes clear is that the architecture can be changed, and that the willingness of doctors to care for those who need them most depends on it. As China refines its DRG rollout, the words of its own clinicians — more work means less pay — stand as both a warning and a blueprint for reform.

Subject of Research: Physicians' patient selection decision-making under China's DRG hospital payment reform in Nanjing

Article Title: “More work means less pay”: a qualitative understanding of physicians’ patient selection decision-making under DRG reform in Nanjing, China

Article References: Zhu, X., Strachan, D. L., Tang, S., Marthias, T., Mahal, A., & McPake, B. (2026). “More work means less pay”: a qualitative understanding of physicians’ patient selection decision-making under DRG reform in Nanjing, China. International Journal for Equity in Health. https://doi.org/10.1186/s12939-026-03044-1

Image Credits: AI Generated

DOI: 10.1186/s12939-026-03044-1

Keywords: DRG payment, cream skimming, patient selection, physician behaviour, health policy, health equity, China, Nanjing, financial incentives, hospital payment reform, qualitative research, health insurance

Helicobacter pylori remains one of the most consequential human pathogens on the planet, colonizing the gastric mucosa of roughly half the world’s population and serving as the dominant risk factor for peptic ulcer disease, chronic gastritis, and gastric adenocarcinoma. Eradicating the bacterium has become progressively harder over the past two decades as clarithromycin and levofloxacin resistance has spread, pushing clinicians toward bismuth quadruple therapy and, more recently, toward dual therapies that pair a potent acid suppressant with high-dose amoxicillin. The logic of these dual regimens is elegant: by driving intragastric pH sharply upward, the acid blocker undermines H. pylori’s ecological niche and enhances amoxicillin’s stability and activity against actively dividing bacteria, while restricting the antibiotic arsenal to a single agent helps preserve susceptibility of other organisms. The arrival of vonoprazan, which binds reversibly to the gastric H+/K+-ATPase at a potassium-binding site and achieves faster, stronger, and more sustained acid inhibition than classical proton pump inhibitors, has energized this strategy, but direct comparisons with optimized proton pump inhibitor dual regimens have remained limited, particularly regarding collateral effects on the gut microbiome.

To address that gap, the research team enrolled a total of 110 H. pylori-positive patients and non-randomly assigned them to one of two fourteen-day-style dual regimens. Fifty patients received the P-CAB regimen, consisting of vonoprazan 20 milligrams twice daily combined with amoxicillin 750 milligrams four times daily, while sixty patients received the PPI regimen, comprising esomeprazole 20 milligrams four times daily with the same amoxicillin dosing. Twenty-two H. pylori-negative individuals served as healthy controls for microbiota benchmarking. Eradication success was determined four weeks after completion of therapy using the 13C-urea breath test, a noninvasive assay that exploits the bacterium’s urease enzyme to detect active infection. The investigators also gathered fecal samples at three time points, baseline before treatment, roughly two weeks into the post-therapy window, and at week six, to characterize shifts in microbial diversity, community composition, and predicted functional pathways through 16S rRNA gene sequencing.

On the pharmacological front, the two acid suppressants differ in ways that matter clinically. Proton pump inhibitors such as esomeprazole are acid-activated prodrugs that require an acidic environment to engage the pump irreversibly, meaning their effect accumulates over several doses and is blunted in the very hypersecretory or proton-pump-dense patients who most need acid control. Vonoprazan, by contrast, is active at neutral pH, accumulates in parietal cells, and exerts dose-dependent, reversible inhibition that reaches near-anacidic gastric conditions within hours and persists through once- or twice-daily dosing. That kinetic advantage is thought to explain why vonoprazan-based dual therapy performs well even in regions with high amoxicillin resistance, since sustained acid suppression amplifies amoxicillin’s bactericidal window. The esomeprazole arm in this study used a four-times-daily schedule specifically to maximize acid suppression and keep the comparison fair against the twice-daily vonoprazan regimen.

The headline efficacy finding favored vonoprazan across every analysis population. The P-CAB group achieved an eradication rate of 97.73 percent in the per-protocol analysis, compared with 84.20 percent in the esomeprazole group, a difference that reached statistical significance at P equals 0.04. Although the intention-to-treat and modified intention-to-treat analyses showed the same directional advantage for the vonoprazan regimen, those comparisons did not cross the threshold of statistical significance, a nuance the authors attribute in part to the non-randomized design and the modest sample size. Still, a cure rate approaching 98 percent in patients who completed the prescribed course places the vonoprazan dual approach among the most effective antibiotic-sparing strategies reported, comfortably exceeding the 90 percent benchmark that international consensus guidelines set for first-line eradication therapy.

Equally important for practice, the safety and tolerability picture was essentially identical between arms. Adverse reactions, which in dual amoxicillin regimens typically include diarrhea, nausea, taste disturbances, rash, and abdominal discomfort, occurred at comparable frequencies in the two groups, and patient compliance did not differ significantly between the P-CAB and PPI regimens. Neither regimen caused treatment-related events severe enough to disrupt the comparison, suggesting that the pharmacological intensity of vonoprazan did not translate into a clinical tolerability penalty. For gastroenterologists weighing first-line options, this aligns the choice more squarely on efficacy grounds, since the operational burden of four-times-daily amoxicillin dosing applies to both regimens equally.

The microbiota component of the study adds a layer of ecological context that eradication trials rarely capture. Using 16S rRNA gene amplicon sequencing of the V3 and V4 hypervariable regions, the team quantified alpha diversity, assessed community structure with principal co-ordinates analysis, and predicted functional shifts against the Kyoto Encyclopedia of Genes and Genomes reference pathways. Both eradication regimens delivered a discernible shock to the gut ecosystem: microbial diversity declined after treatment, community composition shifted measurably away from baseline and from the healthy control profile, and predicted metabolic pathways were transiently perturbed. Importantly, these disturbances followed a recovery trajectory over the follow-up period, with communities migrating back toward their pretreatment configuration by week six.

The two regimens left subtly different ecological fingerprints. The early decline in microbial diversity was more pronounced in the vonoprazan group, an observation consistent with the deeper and more sustained acid suppression that P-CABs deliver, since gastric acid and its downstream influence on the entire intestinal milieu shape which microbes survive transit into the lower gut. Yet the vonoprazan group also showed a tendency toward faster restoration of functional pathway profiles, hinting that the compositional damage and the functional damage may recover on different clocks. Across both arms, the researchers documented an increase in potential pathogenic bacteria and a corresponding decrease in short-chain fatty acid-producing bacteria after treatment, a pattern that matters because short-chain fatty acids such as butyrate fuel colonocytes, maintain epithelial barrier integrity, and exert anti-inflammatory signaling throughout the body. The fact that these SCFA producers rebounded over the six-week window will reassure clinicians, though the study’s design cannot exclude longer-lasting effects in some individuals.

The authors caution, appropriately, that the study was non-randomized, single-center, and modest in size, so the eradication difference should be interpreted as hypothesis-confirming rather than definitive, and the microbiota findings describe short-term dynamics only. Even so, the report lands at a moment when clinicians worldwide are retooling first-line H. pylori strategies amid rising antibiotic resistance. Vonoprazan-based dual therapy is already endorsed by growing bodies of guideline literature as a first-line option, and this comparison provides practical reassurance on two fronts simultaneously: it beats an optimized esomeprazole dual regimen where it counts, achieving eradication close to the theoretical ceiling while matching that regimen on safety and adherence, and the collateral microbiota damage it inflicts appears to be transient and self-repairing. For a bacterium that infects billions and drives one of the world’s most common cancers, treatment strategies that maximize cure rates while minimizing ecological and resistance costs represent exactly the kind of progress the field has been seeking, and this study offers a candid, quantified look at both sides of that bargain.

Subject of Research: Comparative efficacy of vonoprazan-based versus esomeprazole-based dual therapy for Helicobacter pylori eradication and short-term gut microbiota changes

Article Title: Comparative effects of vonoprazan-based and esomeprazole-based dual therapy on Helicobacter pylori eradication and short-term gut microbiota changes

Article References: Liu, Y., Qian, X., Yao, J., Fei, C., Zhang, Z., & Jiang, Z. (2026). Comparative effects of vonoprazan-based and esomeprazole-based dual therapy on Helicobacter pylori eradication and short-term gut microbiota changes. Gut Pathogens. https://doi.org/10.1186/s13099-026-00867-9

Image Credits: AI Generated

DOI: 10.1186/s13099-026-00867-9

Keywords: Helicobacter pylori, vonoprazan, esomeprazole, dual therapy, eradication, gut microbiota, 16S rRNA sequencing, potassium-competitive acid blocker, proton pump inhibitor, short-chain fatty acids, amoxicillin, 13C-urea breath test