Teleradiology Contracts and Liability: What to Get in Writing

You credential individual radiologists at your hospital before they report on your patients. The same standard applies when the radiologist sits three states away and is employed by someone else. That means names. A contract that promises reporting by qualified radiologists without naming them is a contract you cannot audit, and when a specific report is questioned you will need to identify and stand behind a specific person.

Ask for the same file you would keep for a visiting consultant: qualification certificates, registration with the relevant state medical council or the National Medical Register maintained under the National Medical Commission, evidence that the registration is current, professional indemnity cover with the sum insured, and a declaration of any restriction or pending proceeding. Then verify the registration yourself against the public register rather than accepting a scanned certificate, and re-verify annually.

Decide your position on subspecialty scope and on overseas-based reporters. Some hospitals restrict certain studies to named subspecialists; some are uncomfortable with reporting performed outside India for reasons that are as much about accountability and data location as about competence. Both positions are defensible. What is not defensible is having no position and discovering after an incident that a neuroradiology case was reported by whoever was awake.

Most teleradiology SLAs are written as an average, which is close to useless. An average turnaround of forty minutes is compatible with a study that took four hours, and the four-hour study is the one that will hurt you. Write the SLA as a percentile commitment by study category: for example, a stated proportion of emergency CT reported within a defined number of minutes measured from study availability in the provider's system, with a separate and tighter commitment for a defined critical subset.

Define the clock precisely and agree where it is measured. From study transmission or from receipt at the provider. Whether a delay caused by your own network suspends the clock. What happens when the provider requests clinical information and the ward does not respond. Providers will reasonably want protection against your delays, and you should concede that in exchange for precision, because a vague clock benefits whoever writes the monthly report.

Then agree what a breach produces. Service credits are conventional and often trivial; the more valuable remedies are operational. An escalation contact reachable within a stated number of minutes, a monthly review of every breach with root cause, and a defined threshold of repeated breach that permits termination without penalty. The honest trade-off is that tighter SLAs cost more per study, and a hospital wanting fifteen-minute stroke reporting should expect to pay for a staffing model that supports it rather than negotiating it for free.

Allocating liability by contract does not remove your exposure to the patient. A patient or their family bringing a complaint or a consumer case will typically proceed against the hospital that treated them, and the hospital's contractual right to recover from its provider is a separate matter to be pursued afterwards. Contracts that say the provider bears full liability for reporting errors are worth having, and they should be read as recovery mechanisms rather than as shields.

So insist on three things. That the provider carries professional indemnity of a stated amount, that the named reporting radiologists are covered under it or hold their own, and that the cover is written on a basis that will still respond to a claim brought some years after the report, since medico-legal claims in imaging often surface late. Ask for the certificate annually, not once at signature.

Then handle the clinical governance side, which contracts cannot fix. Reports from external radiologists should be included in your discrepancy or peer review sampling on the same basis as internal ones, and the results should be discussed with the provider. A hospital that has never reviewed a single outsourced report has outsourced the reading and retained the risk, which is the worst available arrangement.

Under the DPDP Act 2023 the hospital is the data fiduciary for its patients' personal data and a teleradiology provider processing that data on your instructions is a processor. The Act requires that processing by a processor happens under a valid contract, and that means specific terms rather than a general confidentiality clause: the purpose, the security measures, whether sub-processing is permitted and to whom, what happens on termination, and the provider's obligation to assist you when a breach occurs or a patient exercises a right.

Decide what actually needs to travel. Full patient demographics are not always required for a report, and some hospitals send a reduced identifier set with the clinical history that the radiologist genuinely needs. This is a real trade-off: reduce too much and you degrade the report, because a radiologist reading without age, sex and clinical context is being set up to fail. The workable middle is to send what is clinically necessary and nothing beyond it, and to write down what that set is.

Then cover the mechanics. Transmission encrypted in transit, storage encrypted at rest, access limited to named individuals with individual logins rather than a shared departmental account, retention at the provider limited to a defined period after report delivery, and a documented deletion process. Ask where the data physically sits, and get the answer in writing. HealUDoc activity logs can record which studies were transmitted, when, and under which arrangement, so the disclosure trail exists on your side and not only on the provider's.

Accreditation standards treat outsourced clinical services as your responsibility, not as something you have delegated away. The expectation is that the service is covered by a written agreement, that the provider's competence was verified before the arrangement began, that performance is monitored against defined criteria, and that the monitoring is reviewed by hospital management with recorded outcomes. Your scope of services document also has to state that this examination or this reporting session is provided by an outsourced arrangement.

In practice the artefacts an assessor asks for are predictable: the agreement itself, the credentialing files for named reporters, the monthly performance data against the SLA, evidence that breaches were reviewed, and evidence that report quality was assessed and not only report timing. That last one is where hospitals are thinnest. Timing data comes automatically from systems; quality data requires someone to look at reports.

Also make sure your patients and referring clinicians can tell where a report came from. A report signed by an external radiologist should identify the reporting doctor and their registration, and the referring physician should have a route to discuss the case with them. Anonymous reports arriving from a service with no way to ask a question are a clinical communication failure regardless of their accuracy.

Contracts are negotiated with energy and then filed. The teleradiology arrangements that work well have a standing monthly review with a fixed agenda: volume by category, turnaround against SLA by percentile, breaches with root cause, critical findings communicated and acknowledged, discrepancies identified through peer review, credentialing changes, and any incidents. Half an hour, with the provider on the call, minuted. Without it, the first serious conversation you have with your provider will be during a crisis.

Keep the named reporter list current, because it changes without anyone telling you. Ask for it quarterly and reconcile it against the signatures actually appearing on reports in your system. A name appearing on reports that is not on your credentialing list is an immediate escalation, and it is a check almost nobody runs.

Finally, plan for exit before you need it. Notice periods, transition assistance, return and deletion of data, and a realistic view of how you would cover overnight reporting for the sixty days after termination. Hospitals become operationally dependent on teleradiology faster than they expect, and a provider who knows you have no alternative is negotiating from a position you gave them. Maintaining even a thin internal on-call capability is expensive and is the cheapest insurance in the arrangement.