← Toutes les ressources
Operations · 8 min

The 7 healthcare back-office KPIs nobody tracks (and should)

Billing delay, primary rejection rate, administrative cost per act: the indicators that change decisions.

Most operators steer clinical activity tightly but leave the back-office in a gray zone. They know procedures per practitioner, room utilization, patient satisfaction. They often ignore average billing delay, administrative cost per procedure, or primary rejection rate per fund.

Seven KPIs are enough to regain control. First indicator: average billing delay, measured between procedure date and transmission date. A delay above 5 days signals a bottleneck to investigate.

Second: primary rejection rate, separated public and private payers. A rate above 8% on public or 15% on private signals an upstream billing quality problem, not just a payer problem.

Third: follow-up rate. How many invoices require follow-up before payment? A high rate indicates fragility in the transmission chain.

Fourth: patient DSO. The reference indicator for the patient share, which keeps growing as private insurance coverage decreases.

Fifth: administrative cost per procedure. Calculated by dividing administrative payroll by billed volume, it allows comparing operational efficiency between sites and over time.

Sixth: automation rate. What share of repetitive tasks runs without human intervention? A low rate signals an automation opportunity. A high rate without quality counterpart signals over-investment.

Seventh: administrative team satisfaction. Measured by light quarterly survey, it predicts turnover and work quality. An exhausted administrative team makes more errors and costs more in recruitment and training.

Together these seven indicators tell a clear story of back-office performance. They allow arbitrating between hires, automation, process redesign and tool change with objective data.

Granit provides the standard dashboard to instantiate in 48 hours on existing data, without heavy IT project. Most operators discover on this occasion gaps between sites they did not suspect.