Product details
Reports, dashboards and answers
What your reviewers see on every claim, what managers follow across the portfolio, and what IT and risk teams usually ask.
On every claim
- System results
- Pre-claim anticipation, claim reduction probability and estimate, price anomaly detection, expense profile, priority and severity indices.
- Rule-based profile
- Every rule shown as passed or failed.
- Claim reports
- Cost analysis, consumption analysis and contract check, with the invoices and price lists one click away.
- Context
- Claim and hospitalisation details, diagnosis and procedures, coverage breakdown and historic data for the same patient or provider.
- Review record
- Investigator findings (abuse, waste), notes, evidence and uploaded files; the manager accepts or rejects.
Dashboard: provider utilisation
For administrators and managers: claims, utilisation, cost distribution and claim trends across providers.
- Filters by year and contract type update every tile and chart
- Totals: submitted, outstanding, approved and rejected claims; paid and claimed amounts
- KPIs including savings achieved through claim audits, claims per insured person, average length of stay and average claim lifetime
- Charts: claim rate over time, incident and accommodation types, top ICD10 codes by cost, most costly and most frequent MDCs
Reports library
Available to every role. Dynamic reports respond to filters; static reports are fixed analyses.
Claims & cost
Healthcare Claims Overview Dynamic
Submitted claims, approval status, TPA predicted vs historical cost by clinic and provider, share of flagged claims.
Hospital Report Static
Cost vs length of stay; case volume vs rejections.
Statistical Monitoring Dynamic
Control charts per provider: predicted and actual average cost, approved share after cuts, claims with no cuts, time to settlement, length of stay.
Providers & doctors
Healthcare Providers Dynamic
Claims by hospital clinic, current vs previous year; medical expenses per category and subcategory per hospital.
Claims per Specialty & Department Dynamic
Top doctors by specialty and by clinic.
Doctor Performance Static
Doctor performance across different variables.
Clinical coding
ICD10 Analytics Dynamic
Most frequent admission and most costly discharge categories, admission-to-discharge transitions, claim cuts by provider.
ICD10 Transition Report Static
Rare high-cost transitions by hospital, with heatmaps of the most frequent suspicious transitions.
Top ICD10 Report Static
Top ICD10 transition outliers.
KEN Analytics Dynamic
Admission-to-discharge transition probabilities across KEN categories.
CCSR Analysis Static
Expense distribution and bootstrap analysis.
Clinical Reports Dynamic
Most costly and most frequent major diagnostic categories (MDC).
Forecasting
Hospital Forecast Static
Monthly hospitalisations with a non-parametric forecast, the relative influence of key factors with SHAP explanations, provider performance (cost, duration, volume) and Pareto analyses of CCSR categories.
Frequently asked questions
Does COASSESS decide whether a claim is paid?
No. COASSESS flags anomalies and ranks claims for review. Investigators record their findings and managers accept or reject the claim.
Does it detect fraud?
It detects anomalies: overcharging, abnormal prices, unusual cost structures, failed rules and unusual connections. Whether a case is abuse or waste is decided by your investigators.
What data does it work with?
Hospital admissions and claims, including invoices, price lists, diagnosis and procedure codes, and policy and contract information. The exact data requirements are defined together in the first step of implementation.
How are the results explained?
Each claim shows which rules passed or failed, with cost, consumption and contract-check reports. Forecast reporting uses SHAP values to show which factors drive the prediction.
Where is it deployed?
In your environment. COASSESS is a containerised application, so it adapts to your infrastructure and its constraints, with API connections to your existing systems and models trained on your data.
How long does implementation take?
It depends on your data and environment. Implementation follows five steps: data requirements, data validation (which also produces an FWA report with insights from your data), model training, deployment and user training.
Who can see what?
Role-based access control with four roles: Administrator, Manager, Investigator and Guest. Investigators see only the claims assigned to them; guests have limited, read-only access.
See it on your own claims
Tell us how your team reviews claims today, and we’ll show how COASSESS fits in.