Skip to content

Hospital BI implementation: stages, data and team

Technology is the most predictable part. What sets the timeline of a BI implementation is defining the rules and validating the numbers with the hospital.

Clic Health teamUpdated 3 min read

Hospital BI implementation is often seen as a technology project, but the part that weighs most on the timeline is management: defining the KPI rules and validating the numbers with people who know the hospital. This article covers the stages, the data needed and who needs to take part.

The stages

  1. Technical conversation: understanding the hospital information system, infrastructure, network and security rules.
  2. Data access: IT creates a read-only user on the information system database and prepares the server for the read agent.
  3. First load: reading the history, usually the last few years, so dashboards already compare with the previous year.
  4. Data preparation: applying the hospital's management rules (income statement structure, cost centers, cost groups, payers).
  5. Validation: comparing dashboards with reports the hospital already knows, until the numbers match.
  6. Go-live: creating users, deciding who sees each dashboard and presenting to the teams.
  7. Routine: scheduled refreshes, follow-up and adjustments.

The data needed

TopicData
Patient careEncounters, admissions, discharges, transfers, departments, beds
BillingPatient accounts and items, payers, account status, denials
FinanceReceivables and payables, collections, payments
Accounting and costsAccounting entries, chart of accounts, cost centers
SuppliesSupplies and drug consumption and dispensing
ManagementManagement income statement structure, cost groups, budget, targets

Management data is usually not in the system and is defined with controllership, often in standardized spreadsheets.

Who takes part from the hospital

  • IT: read-only user, server, network and security rules.
  • Controllership or finance: KPI rules, income statement structure and number validation.
  • Billing: validating accounts, denials and deadlines.
  • Executive board: setting priorities, deciding who accesses what and sponsoring the project.
  • Data protection officer (DPO): assessing data processing. See LGPD and hospital BI.

Validation: where the project is won or lost

Validation is the stage that decides whether the BI will be used. The rule is to compare each KPI with a report the hospital already trusts, such as the month's billing or the trial balance, and explain every difference. The most common causes are the date criterion (accrual or billing date), payers or departments included, account classification and entries outside the period.

A good practice is to run BI in parallel with current reports for at least one close. The article Spreadsheets or BI discusses this transition.

Monthly revenue and result in Clic Health: one of the KPIs validated against the hospital's billing and income statement during implementation. Sample data

What delays an implementation

  • A read-only user or server that takes long to be ready.
  • KPI rules that were never written down and must be decided during the project.
  • No one in controllership with time to validate.
  • Chart of accounts or cost centers changing during implementation.
  • Trying every dashboard at once instead of starting with the board's.

How it works in Clic Health

In Clic Health, the read agent is DataClic, which reads the Tasy database with a read-only user. The management rules and dashboards already exist and are adjusted to the hospital's structure. Details are in Tasy integration and data refresh and the implementation FAQ.

Related questions

How long does a hospital BI implementation take?

It depends on data volume, the hospital's rules and team availability. The technical stage is usually the most predictable; number validation varies the most.

Does anything need to stop at the hospital during implementation?

No. Reading data does not interfere with the information system's operation.

How much history should be loaded?

Enough to compare with the previous year and see trends, usually at least two full years, if the data is consistent.

Clic Health team

Written by the team behind Clic Health: accountants, controllers and developers with more than 20 years of work in hospital management.

Talk to us