Frequency vs Severity

The most expensive tenth of hospital claims on Loop's broker book — 7,400-odd admissions out of 74,339 — absorbed 47.6% of the ₹320.6 crore the policy paid out over the latest twelve months. The cheapest half of all claims absorbed under 7%. A group-health policy is two cost problems wearing one premium: a large volume of small, frequent, predictable claims, and a small volume of large, rare, catastrophic ones. The frequency-versus-severity matrix is the tool that separates them — and the separation is where preventive strategy and catastrophic-cover strategy stop being the same conversation.

This page reads the claims book on two axes at once. The horizontal axis is frequency: how often a category of admission shows up, measured as its share of all claims. The vertical axis is severity: how much each admission costs, measured as the average incurred per claim. Plotted together, the ailment categories sort into corners that don't behave alike — and the corner a category lands in tells you which lever moves it.

The bill is split almost evenly between everyday volume and the catastrophic tail#

The clearest way to see the two-cost-problem structure is to rank every claim by what it cost and ask how concentrated the spend is.

Just over one claim in ten carries nearly half the rupees. The bottom half of claims — 37,000 admissions — together cost less than the most expensive 5%. This is the catastrophic tail, and it is the part of the bill that catastrophic cover, network access, and pre-authorisation discipline exist to manage. The other half of the bill is the opposite shape: a high-volume floor of routine, low-ticket admissions where the lever isn't cover, it's prevention and primary care keeping the body out of the hospital in the first place.

The categories sort into four corners, and three of them are populated#

Crossing frequency against severity puts each ailment category in one of four quadrants. Three of them carry real weight.

The corners describe different medicine.

Rolled up, the three populated quadrants carry almost the entire bill — and the rollup makes the strategic point cleanly.

Half of all claims sit in the everyday-churn quadrant and produce ₹105.9 Cr. One in eight claims sits in the catastrophic tail and produces ₹64.4 Cr. The everyday quadrant moves the bill through sheer volume at a low ticket; the catastrophic quadrant moves it through a high ticket at low volume. The ticket gap is 2.6× — ₹73,688 a claim in the tail against ₹28,337 in the churn.

Cardiac is the most severe high-volume line, and it self-selects into the network#

The catastrophic tail is anchored by cardiac admissions: 3.3% of all claims, ₹1.10 lakh each, ₹26.8 Cr of incurred. No other category with meaningful volume comes close on ticket size. Cancer and oncology follow at ₹65,437 a claim, respiratory at ₹54,959. These three categories are the engine of the severity axis — rare enough that no single member expects them, expensive enough that each one reshapes a quarter's loss ratio.

Where these claims go for treatment is itself a finding. The catastrophic-tail categories are the ones most likely to be settled cashless — paid directly to a network hospital at discharge — rather than reimbursed after the fact.

The pattern is monotonic with severity. The bigger and more complex the admission, the more likely it is planned, pre-authorised, and routed to a network hospital where cashless settlement is possible. Cardiac and cancer admissions clear 62%. The everyday-churn categories drift toward reimbursement: routine infection treated at the nearest hospital, maternity at the family's hospital of choice, general health services almost entirely out-of-network. The severity axis and the cashless axis are reading the same underlying thing — the gravity of the event, and the planning that goes with it.

Maternity is the one common category that also carries a real ticket#

Most categories are clearly one thing or the other — common and cheap, or rare and expensive. Maternity is the exception that sits on the line. It is the third-most-frequent category on the book at 10.6% of all claims, and its average ticket — ₹42,816 — lands almost exactly on the policy-wide average of ₹43,129. It is high-frequency and mid-cost at once: not the trivial ticket of a routine infection, not the catastrophic ticket of a cardiac admission, but a real, repeating, planned expense at scale.

This places maternity in a category of its own on the matrix. The everyday-churn quadrant is cheap per event. The catastrophic tail is rare per member. Maternity is neither — it is the rare combination of common and not cheap. At 10.6% of all claims and ₹33.6 Cr of incurred, it is the most predictable large line on the book, because the population that drives it is identifiable in advance. The Maternity and Delivery page takes the line apart in full.

The relationship mix confirms it. Maternity admissions are 77% spouse-layer and 22% employee-layer — overwhelmingly the working-age dependent population. Unlike the catastrophic tail, which lands on the parent layer (cardiac is 64% parent-driven, cancer 61%), maternity is a young-cohort, planned event. It is the category where the policy can see the cost coming and design around it: defined networks, package pricing, structured pre- and post-natal pathways.

The frequency axis maps onto the dependent layers, the severity axis onto the parents#

The two axes of the matrix line up with the household structure the companion Family Claims page sets out. The everyday-churn categories are spread across the working-age and child layers; the catastrophic-tail categories concentrate sharply on the parents.

The catastrophic tail is, in large part, the parent layer's tail. Nearly two-thirds of cardiac admissions and three in five cancer admissions come off the parents and parents-in-law on the policy — the oldest, sickest layer of the covered household. The everyday-churn categories belong to the working-age layers: employee infections, spouse maternity, child fevers. The matrix and the family structure are describing the same population from two angles. The severity axis is, demographically, the parent axis.

What the matrix changes about how the policy is managed#

The single most common mistake in reading a claims book is to rank categories by how often they appear and call the top of that list the cost problem. By claim count, the top of the book is general health services, routine infections, and maternity — and one of those three costs ₹2,910 a claim. The frequency ranking and the rupee ranking are different lists, set out one each on the Top Categories by Frequency and Top Categories by Incurred pages. Musculoskeletal injury is the largest rupee line on the book and only the seventh-most-frequent category. Cardiac is the most severe high-volume line and the eighth-most-frequent. The matrix exists to stop the count from masquerading as the cost.

Open questions for WHI 2027#