Karen S. Palmer is an expert advisor with EvidenceNetwork.ca and Adjunct Professor at Simon Fraser University. Gordon Guyatt is a Distinguished Professor at McMaster University.
Most Canadians probably don't realize that health care in Canada is quietly undergoing a major transformation in funding that could significantly impact patients. Three provinces – Quebec, Ontario and British Columbia – are implementing a new funding model for hospitals and other provinces are watching with interest.
Canadian hospitals have been traditionally funded through annual lump-sum payments – global budgets – meant to pay for all care each institution delivers. The good thing about global budgets is that they are predictable, stable and administratively simple. The problem with global budgets, critics argue, is they lack incentives to boost efficiency, are not transparent and funding is not targeted to priority areas.
Enter activity-based funding (ABF). Under ABF, hospitals receive a pre-determined fee for each episode of care, intended to fund the bundle of services provided to each patient with a particular diagnosis, regardless of the actual costs for any particular patient. The fee is expected to account for the anticipated complexity, type, volume and intensity of care ordinarily provided to clinically similar patients.
ABF has captured the imagination of some policy-makers as an alternative to global budgets. But is ABF the best way to fund hospitals in Canada? And how will it affect patients?
We recently published a systematic review summarizing the global evidence on ABF from the last 30 years. We found that despite its long history, the impact of ABF remains uncertain. Canada should tread cautiously.
High quality systematic reviews are the gold standard in appraising benefits and harms of health interventions. We screened 16,565 studies looking for those that assessed how ABF affected patients and health care systems compared to other funding mechanisms. Results of our review suggest that ABF is associated with a 24 per cent relative increase in patients discharged from hospital to post-acute care services. We also saw possible increases in readmission to hospital.
Shortening hospital stays is a worthwhile policy objective and may promote patient well-being. After all, most people prefer being home to being in hospital. But patients might require readmission to hospital if discharged too soon. Almost certainly, many will need some level of post-acute care after being discharged from hospital "sicker and quicker."
There's the rub. Although Canada has publicly-funded hospital and physician care, funding for home care, rehabilitation care or other forms of post-acute care in the community is a mixed public-private enterprise. Increased pressure on post-acute care capacity in communities could seriously undermine equitable access in Canada, unless accompanied by substantial increases in public funding.
Our study results also found a possible increase in severity of illness among patients admitted to hospital using ABF. Does this mean that patients were sicker going into hospital? Maybe or maybe not.
Since ABF tends to adjust hospital compensation for severity of illness, there is a financial incentive to code patients so that they appear as sick as possible – a practice known as 'upcoding.' Upcoding may be appropriate if it legitimately represents more accurate patient classification, but it may also be inappropriate if the intent is only to maximize hospital reimbursement. Either way, upcoding is likely to undermine at least one ABF policy objective: controlling costs.
We found no consistent increase or decrease in patient death rates with ABF nor in the volume of care provided to patients, such as how much care they received in hospital or the number of patients treated, though results varied widely across studies.
The inconsistency of results across studies for most outcomes suggests there may be contexts in which ABF is more – or less – likely to harm. The question arises: under what circumstances does ABF lead to better or worse outcomes? Does it have to do with specific attributes of how ABF is implemented? Or particular features of a health care system? Unfortunately, no credible explanations for the differences emerged from the evidence.
So, is ABF a good idea for Canadian hospitals? It depends partly on what we hope to achieve, but also on what risks we are willing to take. Governments may not get the benefits they expect with ABF (though they might), and there may be adverse consequences for which they are unprepared (though there might not be).
A few things seem likely: Governments considering ABF should plan for a probable increase in demand for post-acute care services following hospitalization, be aware of the large uncertainty around impacts on other critical outcomes, and be prepared to monitor and evaluate the impact of ABF.
Implementing ABF in Canadian hospitals remains a leap of faith.