Basic proposal
For every nation that supports its citizens in obtaining health services there are financial limitations that make some care unavailable. Here I propose a system that mediates optimally between our deepest values and the realities of illness and scarcity. It would cut paperwork while spurring and even redirecting medical research.
1. A percentage of the average cost of each form of care, one that is unique to that care, would be provided by a system that incorporated into a basic federal plan the supplemental plans of other agencies (federal and state), of employers and individuals. Greater percentages would funnel money towards care offering the most promise in terms of both need and effectiveness.
2. The calculus would be centered, not on characteristics of the patient, but rather on the condition and the promise offered by a diagnosis and treatment. Hence it would be independent of a person’s prior existing conditions or propensities whether medical, behavioral or financial. Payment percentages would be the same in treating a disfigurement, whether congenital, pathological or accidental (there would be no other cosmetic coverage); lung cancer, whether from smoking or from secondhand smoke; and hypothermia, whether the patient slept in a box, a mansion or a crib.
3. The base percentages would be dynamic, changing with medical innovation, assessments of medical realities, average costs of each form of care, and standards of importance. They would further depend dynamically on what funds remain for the basic federal plan’s fiscal year, and lend their structure to all supplementary insurers.
4. Those governing would periodically reassess and possibly reset available funding and said standards of importance along with the values of two percentages: one above which coverage would be lifted to 100%; and the other below which, cut to nothing at all.
5. A natural division of medical care exists: checking for and treating medical conditions. Here “checking” will comprise of interviewing, examining, testing and even exploratory surgery; and “treating,” of counseling, therapy, medication, hospitalization, corrective surgery, prostheses etc.
6. The importance attached to checking will be derived from its reliability and the likelihood and importance of each indicated follow up treatment.
7. For simplicity a care’s importance will be considered to be independent of whichever purpose brought its recommendation.
8. The current paper trail costs billions, but it is simply about confirming that care had been provided, determining payment responsibility, and assessing prior existing conditions. By this, secondary insurers would no longer get involved in the last two responsibilities. Although pre-existing conditions would not affect coverage, medical history and other factors need to be known by physicians to decide the appropriateness of care; by those evaluating various forms of care to bring accuracy to this approach; and by forensic accountants to detect fraud.
In another article, A Tale of Three Databases, I describe a data complex that could contain such private information while guarding patient anonymity. To keep it confidential, accounting information for all insurers may also be kept at the data complex. That would include previewing for a patient the amount to be billed; and upon the application of care, revealing the patient charge and debiting all underwriters.
9. Medical innovations (or revelations) would update the expectation factors, and by this approach be amplified and placed before us. The coverage for a check would increase with its reliability at diagnosing and the importance of such diagnoses in toto; that for a treatment, with its likelihood of:
· Benefiting biological function.
· Alleviating pain or discomfort.
· Prolonging life.
· Supporting prevention.
· Reducing contagion: Currently there is a contagious disease whose spread is often within a hospital's walls. No one wants to pay for the expense of testing those being hospitalized. This plan may well pick up that expense in its entirety.
· Bringing understanding about a particular ailment.
Difficulties with alternatives
1. Will a patient who has been diagnosed with some ailment accept treatment? Leaving out this likelihood of accepting treatment would over emphasize the values of both checking care in general and checks pointing to unpopular procedures in particular, but keeping track may not be worth the trouble. I will include such a likelihood λ on the chance that it really is worth the trouble. Alternatively in accepting diagnosis, one might be required to contract for either the treatment or partial payment of that diagnosis.
2. The annual health care basic allotment (AHCBA) may become inadequate either by poor planning or the drain of a Katrina-like disaster. There are different ways in which the federal government might address this: allowing patient support to diminish toward the end of the year; reserving a portion of the AHCBA and then moderating inadequacy; or voting a lump increase to that year’s AHCBA.
3. How might supplementary insurers respond to the above?
They could offer and charge for several types of insurance, each determined at issuance:
· Policy A’s payout would track the AHCBA as originally budgeted and available.
· Policy B’s payout would track the AHCBA adjusted to include availability increases.
· Policy C’s payout would track the AHCBA adjusted to include both availability and budget increases.
· Policy D’s payout would track a theoretical AHCBA whose remaining available budget was always on target with the projected funding for the available portion of that year’s allotment.
· Policies X+ n% would act as X (e.g. A, B, C, D) while boosting all payments by n%.
4. When non-governmental insurance policies are issued the financial interests of the issuer, those with and those without pre-existing conditions are in conflict with one another. If there is more than one such issuer, then in a system that allows no such distinctions as to payout, the one offering the best deal to those with pre-existing conditions or propensities would soon be bankrupted.
There seem to be at least these possibilities:
· Let each issuer of insurance charge by its own actuarial tables but payout by the policy types (A, B, C, D etc). Those with a pre-existing conditions would be the big losers as well as anyone concerned with the privacy of their health information. However, because of its appeal to the healthy, if allowed it would dominate the market for individuals and their families.
· Establish a futures market for bundled policies of each type. The public would buy the policies it wants (restricted perhaps by current type of coverage). Each policy would have an associated age of insured, proposed portion of per capita AHCBA, and the average number of available but uninsured months leading up to that portion. From this and for each month, enormous bundles of contracts (representing the types and of various statistical deviations from the greater population) could be sliced up and traded on a futures market.
Depending upon how much a bundle were allowed to deviate from a normal profile, this could make such insurance less available to the elderly, to those seeking greater coverage, and to those coming in after staying out for whatever reason: all profiling a pre-existing condition but capturing those quite healthy.
5. The basic percentages will rely on thousands of educated guesses and will be in continuous flux. In their many small adjustments there will be an overall stability in support but no more fixed amounts. It is fanciful of me to imagine that several nations might take an interest and eventually share these guesses. I envision several datapedias (data encyclopedias) each designed much like wikipedia. Their entries would each have a single value along with a listing of independent pre-guesses (with their derivations) which combined to form that entry. There would be datapedias for check and treatment effectiveness regarding all the criteria listed in part nine of the previous section. (May 6, 2010)
6. There is a subtle conflict between the quality and privacy of data. Those immersed in it may need to report fraud or seek clarification. For the sake of privacy non-identifying information would be kept at one data center with an attached (possibly floating) id number, while a link between a patient’s public identity and id number would be kept under strict supervision at a separate center (see my other article A Tale of Three Databases). In the case of fraud a judge would need to issue a warrant to search out this otherwise private identity.
7. All over the world there are nations taking on the health care of their citizens: each with its own generosity, each with its own triage. This approach offers a range wherein the public might better understand how its survival and wellbeing are written out, and here may lay its major difficulty: being perceived as a branching away from both public and private health plans, when in fact it is to each a natural next step.
Logistics
Up until now I have outlined reconciliation between the realities of illness and the healthcare to which we might aspire. Now I render this in terms that are less magical but that any self-respecting computer should understand.
Data structures for a given moment
1. Let B be the set of all biological functions that spell out the quality and duration of human life. All medical treatments consist of supporting effective function or, when an abnormality threatens death or great pain, removing its physical matrix.
2. Define I: B --> R by I (β) = the importance of biological function β being at 100% of its natural effectiveness. Assignment of I (β) would come out of committee to legislative and executive branches.
3. Let T be the set of all treatments.
4. Let Χ be the set of all checks χ, and subset б (χ) of C contain that care which χ may recommend.
5. Let C be the set of all care, and for c ε C let subset Ъ(c) of B contain those biological functions possibly impacted by the carrying out of c. This includes side effects of treatments and check.
6. For τ ε T define G (τ): Ъ (τ) --> [-100,100] by G (β, τ) = the average gain in the percentage of β’s effectiveness, when treatment τ is undertaken. Assignments of G (τ) would be by specialists in the underlying need.
7. Since there may be dis/synergies in the joining of forms of care into a battery of care, a simple union may not be reliable. Such a battery ought to be dealt with as a meta-care. When both treatment(s) and check(s) are joined the resulting battery will be called hybrid. If a check has positive and negative side effects, it will be dealt with as a hybrid battery whose treatment has those effects.
8. For χ ε Χ let Λ (χ): б (χ) x [T+, F+, T-, F-] --> [0, 1] be the likelihood that χ will (True+, False+) or will not (True-, False-) recommend c. Assignment of Λ (χ) would be by specialists in the underlying need.
9. Define λ: C --> I as the likelihood that, being advised to accept a form of care c, a person will in fact accept it.
10. Define Ĉ: C --> $ by Ĉ (c) = Average amount charged for receiving care c. Ĉ (c) would come from an ongoing medical accounting archive.
Calculations for a given moment and in order
1. Define Ē, Ē+, Ē-: T --> R (the expectations for τ) by
Ē (τ) = ∑ Ъ (τ) G (β, τ)*I (β)
Ē+ (τ) = ∑ Ъ (τ) Max (G (β, τ), 0)* I (β)
Ē- (τ) = ∑ Ъ (τ) Min(G (β, τ),0)* I (β).
2. Define A [T, F+, F-]: Χ --> R (the importance of χ implicit in its recommendations) as
A [T]: Χ --> R by A (χ) =
∑ κ ε б (χ) SQRT (Λ (χ) (κ, T+)* Λ (χ) (κ, T-))* λ (κ) * Ē (κ)
A [F+]: Χ --> R by
A (χ) = ∑ κ ε б (χ) Λ (χ) (κ, F+)* λ (κ) * Ē- (κ)
A [F-]: Χ --> R by
A (χ) = ∑ κ ε б (χ) Λ (χ) (κ, F-)* λ (κ) * Ē+ (κ)
3. Define Ē: Χ --> R (the expectation of χ) by
Ē (χ) = A [T] (χ) + A [F+] (χ) – A [F-] (χ)
4. Define €: C --> $ (the augmented charge of c) by
€ (c) = Ĉ (c) when c ε T
€ (c) = Ĉ (c) + ∑ κ ε б (c) Λ (c) (κ, F+)* Ĉ (κ)
when c ε X
Now the above moment redounds to a fiscal year.
Data structures for a fiscal year
1. Let Y be a fiscal year.
2. Define F: Y --> R as funds remaining and available from the AHCBA at the moment m ε Y. This would be determined by accounting.
3. Define N: C x Y --> I as the projected number of requests for care c from moment m to the end of Y. Unless care c is directed towards a seasonal illness this would likely be the portion of the year remaining times the annual anticipated need (flu being a clear exception).
4. Define %: C x Y --> [0, 100] as the base percentage associated with care c at moment m.
5. Let M+ & M- be those percentages such that above M+ coverage would be 100% and below M-, none at all.
6. Let S be an ordered set of underwriters of supplementary insurance. For s ε S define P(s): USA --> R+ as the relative commitment to someone in the USA as a portion of the basic per capita AHCBA set at issuance.
Calculations
1. Let e: C x Y--> R (effectiveness of a dollar spent on c) be given by
e(c, m) = Ē(c)/ € (c) calculated at the moment m.
2. Define emax: Y --> R (the maximum value of e(c, m) over all forms of care, c).
3. Dollar support should flow in the direction of its greatest effectiveness, thus %( c, m) is proportional to e(c, m) and in fact
%(c, m) = 100* e(c, m) / emax (m).
4. Since F(m) must make it through the year (Y),
F (m) = ∑all careN(c, m) * € (c, m)
5. From a & b, it can be shown that %( c, m) =
100 * F (m) * e(c, m) / ∑all careN(c, m) * Ē(c, m)
6. Let U%, L%: USA x C x Y -> [0,100] be the unlimited and limited total % coverage.
Then U% (x, c, m) =% (c, m) * (1 + ∑ s ε S P(s,x))
And L% (x, c, m) =
100% (for U% (x, c, m) >= M+)
0% (for U% (x, c, m) < M-)
U% (x, c, m) otherwise
For s ε S underwriter s would be responsible for the portion
P(s)/ (1 + ∑ s ε S P(s,x)) of L% (x, c, m)
7. Let Ŧ: USA x C x Y -> $ be the total coverage afforded someone in the USA for care c at moment m.
Then Ŧ (x, c, m) = Ĉ (c) * L% (x, c, m) /100
And the co-pay for person x having received care c at moment m is the provider’s charge - Ŧ (x, c, m).