A 19 percent drop sounds like victory. It might be.
Since 2019, Victoria's SafeScript system has quietly reshaped the landscape of prescription medicine, recording a 19 percent decline in patients obtaining monitored drugs from four or more doctors. Yet this apparent victory invites a deeper question that haunts many well-intentioned interventions: does a measurable change in behavior reflect genuine harm prevented, or merely the displacement of risk and the silencing of legitimate need? The first rigorous examination of SafeScript's impact reveals not a clean answer, but a more honest reckoning with the limits of surveillance as a proxy for safety.
- A 19 percent drop in multiple-prescriber episodes sounds like a public health win — but researchers are not yet certain it represents lives saved rather than prescribers spooked.
- The system cannot distinguish a chronic pain patient navigating a team of specialists from someone illegally stockpiling opioids, and that ambiguity is now driving clinical behavior.
- Low-risk GPs, confronted with software alerts, may be quietly withdrawing care from patients who never posed a danger — a harm that leaves no obvious trace in the data.
- The people SafeScript was designed to stop — determined, high-risk doctor-shoppers — may simply have adapted, while the statistical decline flatters the system's true effectiveness.
- Researchers are now pressing for follow-up studies that can connect prescribing pattern changes to actual outcomes: overdoses, addiction rates, and untreated pain.
Victoria launched SafeScript in 2019 with a clear mandate: intercept patients collecting opioids and benzodiazepines from multiple doctors before addiction or overdose could take hold. Seven years on, the first serious attempt to measure whether it worked has produced a number that looks like success — and a set of questions that complicate it entirely.
A study in the Medical Journal of Australia tracked 562 GP practices and found that multiple-prescriber episodes fell from 15.73 to 12.6 per 1,000 people after the software launched — a 19 percent reduction. On the surface, the system appeared to be doing precisely what it was designed to do.
But a GP and addiction medicine specialist reviewing the findings pressed on what that decline actually represents. The software flags any patient who obtains monitored medicines from four or more prescribers within 90 days. That pattern looks the same whether the patient is a chronic pain sufferer coordinating care across a GP, a pain specialist, and a rheumatologist, or someone deliberately collecting pills. When doctors see the alert, many may simply become more conservative — refusing to prescribe even when it is clinically appropriate. The drop in episodes may reflect not less doctor-shopping, but more cautious prescribing among doctors who were never the problem.
This is the quiet cost embedded in broad-net monitoring systems. They change behavior efficiently, but they cannot easily distinguish between the behavior they were meant to stop and the behavior that was never harmful to begin with. A statistical decline is not the same as harm prevented.
What the research has not yet shown is whether SafeScript actually reduced overdoses, addiction, or death — or whether patients with genuine pain needs are now going without care, or whether high-risk individuals simply found other routes to the drugs they sought. The system may be a genuine safety tool. It may also be a system that looks effective because it moved numbers, not because it moved outcomes. Untangling those two possibilities will require considerably more work.
Victoria introduced SafeScript in 2019 with a straightforward goal: stop patients from doctor-shopping for dangerous drugs. The software was meant to catch people collecting opioids and benzodiazepines from multiple prescribers — the kind of behavior that feeds addiction and overdose. Seven years later, researchers have their first real measure of whether it worked.
A study published in the Medical Journal of Australia examined 562 GP practices and found something that looks like success. The number of patients obtaining monitored medicines from four or more doctors within a 90-day window dropped 19 percent after SafeScript launched. Before the software, the rate sat at 15.73 episodes per 1,000 people. After it went live, that fell to 12.6 per 1,000. On paper, the system appeared to be doing exactly what it was designed to do.
But the story gets more complicated when you ask what that number actually means. A GP and addiction medicine specialist who reviewed the findings raised a crucial question: Did SafeScript stop dangerous doctor-shopping, or did it simply change how low-risk prescribers behave? The distinction matters enormously. A patient with legitimate chronic pain who sees multiple doctors for genuine medical reasons — a pain specialist, a GP, a rheumatologist — might look identical in the data to someone illegally collecting pills. When SafeScript flags these episodes, low-risk prescribers might become more cautious, more reluctant to prescribe, even when the patient's needs are genuine.
This is the hidden cost of prescription monitoring systems. They cast a wide net. They cannot easily distinguish between the patient seeking help and the patient seeking to feed an addiction. The software sees the pattern — four prescribers, 90 days — and raises an alert. Doctors, seeing that alert, may change their prescribing habits. They may become more conservative. They may refuse to prescribe at all, even when it's appropriate. The result is a decline in multiple-prescriber episodes, but not necessarily because doctor-shopping has stopped. It may simply be because doctors have become more afraid.
The study itself is solid work. It is the first research to measure SafeScript's actual impact on prescribing behavior since the system went live in 2019. The researchers had real data from real practices. They could see the before and after. But data alone cannot tell you whether a change is good or bad, whether it reflects genuine harm prevention or unintended consequences. A 19 percent drop sounds like a victory. It might be. Or it might be a sign that the system is working too broadly, catching too many people, changing behavior in ways that harm patients who were never the problem in the first place.
What happens next is unclear. The researchers have documented that SafeScript changed prescribing patterns. What they have not yet shown is whether those changes actually prevented overdoses, addiction, or death. They have not shown whether patients with legitimate pain management needs are now going without treatment. They have not shown whether the people who were actually doctor-shopping simply found other ways to get drugs. The software may be a genuine safety tool. Or it may be a statistical artifact — a system that looks effective because it changed behavior, not because it prevented harm. The difference between those two things will require more research to untangle.
Citações Notáveis
The findings could reflect low-risk prescribers changing their behaviour, rather than a decline in dangerous doctor-shopping— GP and addiction medicine specialist quoted in the study
A Conversa do Hearth Outra perspectiva sobre a história
So SafeScript cut multiple-prescriber episodes by 19 percent. That sounds like it's working.
It does sound that way. The numbers are real — 15.73 down to 12.6 per 1,000 people. But the question is what those numbers actually represent.
What do you mean? Fewer people getting drugs from multiple doctors is fewer people doctor-shopping, right?
Not necessarily. It could also mean that low-risk prescribers — doctors treating patients with legitimate pain — became more cautious after the system flagged their behavior. They might have changed how they prescribe, even though their patients weren't the problem.
So the software caught the wrong people?
It caught everyone. It can't tell the difference between a patient with chronic pain seeing multiple specialists and a patient illegally collecting pills. Both look the same in the data.
And doctors responded by prescribing less?
Possibly. When you flag a behavior, people change it. Whether that change prevents harm or just prevents treatment is the question nobody has answered yet.
So we don't actually know if SafeScript is making people safer?
We know it changed prescribing patterns. We don't know if it prevented overdoses, addiction, or death. That's the gap the research needs to fill.