The expected-value cost of a failed implant, done as math

A two-thousand-dollar saving on an implant quote is real money, and I am not going to wave it away with a lecture about quality. If you can have the same outcome for less, you should, and pretending price does not matter is its own kind of dishonesty. So let me concede the appeal of the cheap quote at full strength. The saving is concrete, immediate and easy to understand.

The problem is that quote-versus-quote is the wrong calculation, and it is wrong in a specific, fixable way. A quote is the price of one outcome: the outcome where everything goes right. But you are not buying one outcome. You are buying a probability distribution over outcomes, most of them good and a few of them expensive. The figure that tells you which option is actually cheaper is not the sticker price. It is the expected value, the probability-weighted total cost across the outcomes that actually occur [1]. This piece builds that calculation from scratch, with clearly labeled illustrative variables and no invented prices, so you can drop in your own numbers and see the answer for your own situation.

Expected value, in one paragraph

Expected value is a weighted average. You list the possible outcomes, multiply each one’s value by its probability, and add them up [1]. The formula for a finite set of outcomes is the sum of each outcome times its probability: E = x1 times p1 plus x2 times p2, and so on. That is the entire idea. It is the honest figure because it does not assume the good outcome; it accounts for every outcome at the rate it occurs. The reason it matters here is that a sticker price quietly assumes the success outcome with probability one, which is never true, and the gap between that assumption and reality is exactly where a cheap quote can hide an expensive bet.

The variables, defined and labeled as illustrative

I will not put fabricated clinic prices in your mouth, because invented numbers masquerading as data are precisely the thing this publication refuses to do. Instead, here are clearly labeled variables. You supply the values from real quotes and defensible probabilities; the framework supplies the structure. Every symbol below is an illustrative placeholder, not a claimed figure.

• Q = the quoted upfront price you pay for the implant treatment.
• p = the probability of success for that treatment, as a decimal (so 0.95 means a 95 percent chance of success).
• f = the probability of failure, which is simply 1 minus p.
• R = the full cost of putting a failure right: the revision. Crucially this is not just re-doing the implant. It includes any bone grafting needed after a failure, a second round of surgery, repeat travel, accommodation, lost work time, and the premium a new clinic may charge to take on someone else’s failed case.
• The expected total cost of an option is then: E = Q plus (f times R).

That last line is the whole tool. The expected cost is what you pay if it works, plus the cost of failure scaled by how often failure happens. Two options should be compared on E, not on Q.

A note on where p comes from, because this is where people either cheat or freeze. Published implant survival figures provide a general anchor: in healthy tissues with appropriate loading, five-year survival is commonly cited in the mid-to-high nineties percent, which corresponds to a low single-digit failure rate as a starting reference [2]. You then adjust for your own situation. Smoking, uncontrolled diabetes, poor bone quality, heavy parafunction, and, critically, an implant whose brand and lot number you could not verify, all push f upward [2]. The point of the framework is not to hand you a precise f. It is to force you to pick a defensible one rather than silently assuming f equals zero, which is what comparing sticker prices does.

The worked framework as a table

The table below shows the structure with illustrative variable values, chosen only to demonstrate the arithmetic. They are labeled placeholders, not real prices or real clinic data. Replace every value with your own and recompute. The comparison is between a cheaper Option A with weaker odds and a costlier path to fixing failures, and a dearer Option B with stronger odds and a cheaper revision path.

Read the bottom three rows. The sticker comparison looks only at Q and declares Option A the winner because Q_A is lower. The expected-value comparison adds the failure term. Option A carries a failure probability more than three times higher in this illustration, and a revision that is more expensive because follow-up is distant and a new clinic must inherit the case. Option B fails rarely and is cheaper to fix when it does. Whether A or B wins on E depends on your actual numbers, but the structural lesson is fixed: the entire failure term is invisible in the sticker comparison and can easily exceed the headline saving. A saving of S on the quote is only a real saving if S is larger than the difference in the two failure terms.

Why the revision cost R is usually the surprise

People underestimate R, and it is worth slowing down on why, because R is where the math turns. A revision is not a discounted redo. After an implant fails, the bone that held it may be compromised, so the next attempt can require grafting and a healing delay before a new implant is even possible [2]. If the original treatment involved travel, the revision often does too, doubling the logistics, and a clinic that did not place the original implant has every reason to price a complex salvage case at a premium. There is also the cost that never appears on an invoice: time off, repeated procedures, and the uncertainty of starting over.

This is also where a separate cognitive error tends to ambush people: the sunk-cost fallacy [4]. Having paid Q_A, a patient feels the money is “in” the cheaper clinic and is tempted to return there to protect the original spend. But money already paid is gone regardless of what you do next, and it should carry zero weight in the forward decision [4]. The only honest question at the point of failure is the expected cost of each path from here, looking forward. The framework is built to look forward, which is exactly why it ignores what you already spent. This is decision theory’s core stance: choose the action with the best expected outcome given current information [3].

How verification changes the probabilities, not just the paperwork

The variable p is not fixed by the universe. It is influenced by choices you can audit before you commit, and this is where the other tools in this series feed directly into the math. An implant whose brand and lot number you could not confirm carries an elevated f, both because unverifiable provenance is associated with elevated counterfeit and quality risk and because a future clinic may struggle to source matching components for a revision. That is the practical stakes behind verifying an implant brand and lot number before surgery. Difficult follow-up raises R, which is part of why the records to obtain before you leave a dental clinic abroad is not bureaucratic box-ticking but a direct input to the revision cost.

And the failure term is exactly the term that the before-and-after photo as a survivorship-bias trap shows you can never read off a gallery, because the gallery contains zero failures by construction. The expected-value model and the survivorship model are two views of the same blind spot: the cases that fail are both the ones missing from the photos and the ones driving the cost. The structural incentives that keep f opaque are the subject of the package-deal overtreatment incentive, and a ready-made arithmetic companion to this piece is the break-even calculator, which runs the same logic against travel and treatment savings.

What a patient should verify

The framework is only as good as the inputs you can defend, so the verification is about pinning down each variable honestly.

• Q: get the full, written, all-inclusive quote, including components, abutments, restoration and any staged costs, so Q is the real total and not a teaser.
• p and f: choose a defensible success probability anchored to published survival data, then adjust for your specific risk factors rather than assuming success is certain [2].
• R: estimate the genuine all-in revision cost, including possible grafting, a second surgery, repeat travel, time off, and the premium of a new clinic inheriting the case.
• Confirm the warranty’s real coverage and exclusions, because a warranty that pays part of R lowers your effective failure cost, and one that excludes the likely failure modes does not.
• Compute E = Q + (f times R) for each option and compare those, never the bare quotes.
• Exclude any money already spent from the forward decision; it is a sunk cost [4].

The reasoning is falsifiable in the strict sense: if, after honestly filling in Q, f and R for two real options, the cheaper quote still has the lower expected total cost, then the cheaper quote genuinely is cheaper and you should take it. The framework does not push you toward expensive care. It pushes you toward the option with the lower expected cost, whichever that turns out to be.

The honest bottom line

The saving on a cheap quote is real, but a quote is the price of the good outcome, and you are not buying only the good outcome. You are buying a distribution, and the few expensive outcomes in the tail are paid for in full by whoever draws them. Expected value is just the discipline of pricing the whole distribution instead of the brochure.

Write down Q, pick an honest f, estimate the true R, and compute Q plus f times R for each option. Compare those numbers. Ignore what you have already spent. Sometimes the cheap quote wins this calculation cleanly, and when it does, take it with confidence. Often it does not, and the saving you were celebrating turns out to be smaller than the probability-weighted cost of the failure nobody quoted you. The math is not cold. It is the only version of the comparison that includes the outcomes you were hoping not to think about.

For the radiation arithmetic of the same shopping process, see cumulative CBCT radiation across a multi-clinic shopping journey. For the logistics decision the same trip dresses up as biology, see why flying home after an implant is set by airfare, not biology. For whether to travel at all, see when to go overseas for dental treatment and the dental tourism trust gap. Our standards are at methodology and disclosures.