Method of diagnosis of lyme disease

A method of determining the probability that a patient has Lyme disease by interpreting the results of an immunoassay which measures individual antibody responses to multiple separate antigens associated with Lyme Disease (L) comprising (a) employing scored clinical criteria to obtain a patient's pretest probability of having the disease, (b) carrying out an immunoassay on the patient's body fluid, and, (c) interpreting the results of the immunoassay through an algorithm wherein each individual antibody response is viewed as a separate test performed in parallel, wherein the posttest probability that the patient has Lyme disease is derived as a consequence of the pretest probability that the patient has Lyme disease and the exact combination of antibody responses as measured by the immunoassay as follows:

• given that i, j, kε of size n, but i≠k, P(L) is the prior probability of Lyme Disease, P(Bj/L) is the probability of the jth immunoassay response, Bj, given Lyme Disease is present;
• Bi is the ith antibody response;
• Bk is the kth antibody response;
• B1B2=B1∩B2, P(Bj)≠0, L∩Bj≠Ø, and P(Bj/L) is significantly greater than P(Bj/L′) (P<0.01 for each jεJ);
• then the post-test probability that the results of a given patient's immunoassay indicate infection with Lyme Disease is determined using the following formula:
• P ⁡ ( B l ⁢ ⁢ … ⁢ ⁢ B j ⁢ B j + 1 ′ ⁢ ⁢ … ⁢ ⁢ B n ′ / L ) × P ⁡ ( L ) P ⁢ ( B l ⁢ ⁢ … ⁢ ⁢ B j ⁢ B j + 1 ′ ⁢ ⁢ … ⁢ ⁢ B n ′ / L ) × P ⁢ ( L ) + [ ∏ i = 1 j ⁢ ⁢ P ⁡ ( B i / L ′ ) ] × [ ∏ k = j + 1 n ⁢ ⁢ P ⁡ ( B k ′ / L ′ ) ] × P ⁡ ( L ′ )
• and wherein:
• B1, . . . , Bn represents a group of antibody responses measured by the immunoassay which are associated with infection with Lyme Disease and measured by the immunoassay test for a given patient;
• B1, . . . , Bj represents those antibody responses included in B1, . . . , Bn which are present in the results of a given immunoassay;
• B′j+1, . . . , B′n represents those antibody responses included in B1, . . . , Bn which are not present in the results of a given immunoassay;
• P(B1 . . . BjB′j+1 . . . B′n/L) represents the frequency that a given immunoassay pattern is present in patients with Lyme Disease;
• P(Bi/L′) represents the frequency with which the antibody response, Bi appears in the immunoassay when there is no infection with Lyme Disease;
• [ ∏ i = 1 j ⁢ ⁢ P ⁡ ( B i / L ′ ) ]
• represents the product of the frequencies with which the individual antibody responses present in the given immunoassay pattern appear in the immunoassay when there is no infection with Lyme Disease;
• P(B′k/L′) represents the frequency with which a antibody response, Bk is absent in the immunoassay when there is no infection with Lyme Disease (i≠k), [ ∏ k = j + 1 n ⁢ ⁢ P ⁡ ( B k ′ / L ′ ) ]
• represents the product of the frequencies with which the individual antibody responses not present in the immunoassay pattern are absent in the immunoassay when there is no infection with Lyme Disease;
• and,
• P(L′) represents a pre-test probability of no infection with Lyme Disease.