Apparatus, method and product for multi-attribute drug comparison

• US 6,529,892 B1
• Filed: 08/04/1999
• Issued: 03/04/2003
• Est. Priority Date: 08/04/1999
• Status: Expired due to Fees

• Current Assignee: Bruce L. Lambert
• Original Assignee: University of Illinois
• Inventors: Lambert, Bruce L.
• Primary Examiner(s): Black, Thomas
• Assistant Examiner(s): Booker, Kelvin
• Application Number:
  US09/368,203
• Time in Patent Office: 1,308 Days
• Field of Search: 706/55, 706/1, 706/52, 706/20, 706/48, 706/45, 707/5, 707/3
• US Class Current: 706/55
• CPC Class Codes: G06F 40/226 Validation
• CPC Class Codes: G06F 40/268 Morphological analysis
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06F 40/295 Named entity recognition
• CPC Class Codes: G16H 20/10 relating to drugs or medications, e.g. for ensuring correct administration to patients
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients
• CPC Class Codes: G16H 70/40 relating to drugs, e.g. their side effects or intended usage
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A computer-implemented apparatus or method, or a software product, for generating a composite quantitative comparison of drug products based on multiple attributes of them. A set of name-attribute similarity scores are generated based on similarities among the names of selected target and reference drugs. A set of product-attribute similarity scores are generated based on similarities among product attributes of the selected target and reference drugs. A target drug confusability score is generated based on the confusability of the target drug as compared to a population of other drugs. The composite quantitative comparison is generated based on a composite of the name-attribute and product-attribute similarity scores, and the target confusability score.

Claims

1. A drug comparator for assessing the confusability of one or more target drugs and one or more reference drugs based on one or more objective attributes of each target drug and each reference drug, comprising:

a product attribute comparator that deterministically generates one or more product-attribute similarity scores each representing a similarity of a product attribute of the one or more target drugs and the one or more reference drugs.

2. The drug comparator of claim 1, further comprising:

a name attribute comparator that deterministically generates one or more name-attribute similarity scores each representing a similarity of a name attribute of the one or more target drugs and the one or more reference drugs.

3. The drug comparator of claim 2, wherein the name attribute comparator comprises:

an orthographic analyzer that generates one or more name-attribute similarity scores based on one or more orthographic measures of the name attributes of the target and reference drugs.

4. The drug comparator of claim 3, wherein the one or more orthographic measures comprises:

one or more N-gram measures.

5. The drug comparator of claim 3, wherein the one or more orthographic measures comprises:

one or more edit distance measures.

6. The drug comparator of claim 2, wherein the name attribute comparator comprises:

a phonetic analyzer that generates one or more name-attribute similarity scores based on one or more phonetic measures of the name attributes of the one or more target and one or more reference drugs.

7. The drug comparator of claim 6, wherein the one or more phonetic measures are determined by phonetically transforming the name attribute of at least one of the target and reference drugs.

8. The drug comparator of claim 7, wherein the one or more phonetic measures are determined by applying an orthographic measure to the phonetically transformed name attribute.

9. The drug comparator of claim 2, wherein the name attribute comparator comprises:

a phonological analyzer that generates one or more name-attribute similarity scores based on one or more phonological measures of the name attributes of the one or more target and reference drugs.

10. The drug comparator of claim 2, wherein the name attribute comparator comprises:

an orthographic analyzer that generates one or more name-attribute similarity scores based on one or more orthographic measures of the name attributes of the one or more target and reference drugs;

a phonetic analyzer that generates one or more name-attribute similarity scores based on one or more phonetic measures of the name attributes of the one or more target and reference drugs; and

a phonological analyzer that generates one or more name-attribute similarity scores based on one or more phonological measures of the name attributes of the one or more target and reference drugs.

11. The drug comparator of claim 2, further comprising:

an attribute similarity scores processor that mathematically combines a plurality of attribute similarity scores comprising the group of attribute similarity scores consisting of the one or more name-attribute similarity scores and the one or more product-attribute similarity scores to generate one or more processed attribute similarity score.

12. The drug comparator of claim 11,

wherein at least one of the plurality of attribute similarity scores is associated with a weight, and

wherein the attribute similarity scores processor applies the weights to their associated attribute similarity scores prior to combining the attribute similarity scores.

13. The drug comparator of claim 12,

wherein the weight of the at least one attribute similarity score is determined by a mathematical/logical function.

14. The drug comparator of claim 12,

wherein the attribute similarity scores processor generates the processed attribute similarity score by determining a weighted sum of the one or more attribute similarity scores.

15. The drug comparator of claim 12, wherein:

the weight of each of the plurality of attribute similarity scores is determined by a user.

16. The drug comparator of claim 11,

wherein the composite score generator generates a quantitative interpretation that compares the processed attribute similarity scores corresponding to a first set of target and reference drugs to a processed attribute similarity scores corresponding to a second set of target and reference drugs.

17. The drug comparator of claim 2,

wherein the one or more name-attribute similarity scores and one or or more product-attribute similarity scores are associated with a weight, and

wherein the drug comparator further comprises an attribute similarity scores processor that applies the weights to their associated attribute similarity scores.

18. The drug comparator of claim 2, wherein, the drug comparator further comprises:

a composite score generator that deterministically generates one or more composite quantitative comparisons based on comparisons of one or more of the product attribute and one or more name-attribute attribute-similarity scores.

19. The drug comparator of claim 1, wherein the product attribute comparator comprises any one or more comparators selected from the group consisting of:

a strength comparator that generates at least a first product-attribute similarity score based on one or more dosage strengths of the target and reference drugs;

an indication comparator that generates at least a second product-attribute similarity score based on one or more indicated uses of the target and reference drugs;

a dosage form comparator that generates at least a third product-attribute similarity score based on one or more dosage forms of the target and reference drugs;

an administration route comparator that generates at least a fourth product-attribute similarity score based on one or more administration routes of the target and reference drugs;

a manufacturer comparator that generates at least a fifth product-attribute similarity score based on one or more manufacturers of the target and reference drugs;

a pharmacological category comparator that generates at least a sixth product-attribute similarity score based on one or more pharmacological categories of the target and reference drugs;

a storage requirements comparator that generates at least a seventh product-attribute similarity score based on one or more storage requirements of the target and reference drugs;

a color comparator that generates at least an eighth product-attribute similarity score based on one or more colors of the target and reference drugs;

a shape comparator that generates at least a ninth product-attribute similarity score based on one or more shapes of the target and reference drugs;

a legal standing comparator that generates at least a tenth product-attribute similarity score based on one or more legal standings of the target and reference drugs; and

a trademark description comparator that generates at least an eleventh product-attribute similarity score based on one or more goods and services trademark descriptions of the target and reference drugs.

20. The drug comparator of claim 1, further comprising:

a neighborhood density calculator that calculates a neighborhood density score for at least a first target drug of the one or more target drugs based on a number of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is a function of the one or more attribute similarity scores.

21. The drug comparator of claim 1, further comprising:

a neighborhood familiarity calculator that calculates a neighborhood familiarity score for at least a first target drug of the one or more target drugs based on a familiarity of the first target drug and the familiarity of each of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is a function of one or more name- or product-attribute similarity scores.

22. The drug comparator of claim 12, wherein:

the neighborhood familiarity score is based on a name of the first target drug and names of the first set of reference drugs.

23. The drug comparator of claim 22, wherein the first measure of familiarity is determined by using frequency of prescription as a proxy for familiarity.

24. The method of claim 21,

wherein the step (4) of combining the attribute similarity scores comprises determining a weighted sum of the one or more attribute similarity scores.

25. The drug comparator of claim 1, further comprising:

a severity of confusion scores generator that generates one or more scores for at least one of the one or more target drugs, each severity of confusion score representing the severity of the consequence to the patient of confusing a target drug for a reference drug.

26. The drug comparator of claim 25,

wherein, the drug comparator further comprises a composite score generator that deterministically generates one or more severity-weighted similarity scores determined by weighting one or more of the attribute similarity scores by one or more of the severity of confusion scores.

27. The drug comparator of claim 26,

wherein the composite score generator further generates a quantitative interpretation that comprises a first set of one or more severity-weighted similarity scores corresponding to a first set of target and reference drugs and a second set of one or more severity-weighted similarity scores corresponding to at least a second set of target and reference drugs.

28. The drug comparator of claim 27,

wherein the reference drugs of the first and second sets are the same.

29. The drug comparator of claim 25, wherein the severity of confusion scores generator comprises:

a therapeutic index comparator that generates at least a first of the one or more severity of confusion scores based on a therapeutic index of one or more of the target drugs and a therapeutic index of one or more of the reference drugs.

30. The drug comparator of claim 25,

wherein the severity of confusion scores generator comprises a therapeutic index comparator that generates at least a first of the one or more severity of confusion scores based, on a contraindication index of one or more of the target drugs and a contraindication index of one or more of the reference drugs.

31. The drug comparator of claim 1,

wherein when a user selects the one or more target drugs, the drug comparator automatically generates the attribute similarity scores.

32. The drug comparator of claim 1, further comprising:

an attribute similarity scores processor that combines a plurality of the one or more product-attribute similarity scores to generate a processed attribute similarity score.

33. The drug comparator of claim 32,

wherein at least one of the plurality of attribute similarity scores is associated with a weight, and

wherein the attribute similarity scores processor applies the weights to their associated attribute similarity scores prior to combining the attribute similarity scores.

34. The drug comparator of claim 33,

wherein the attribute similarity scores processor generates the processed attribute similarity score by determining a weighted sum of the one or more attribute similarity scores.

35. The drug comparator of claim 33,

wherein the weight of the at least one attribute similarity score is determined by a mathematical/logical function.

36. The drug comparator of claim 33, wherein:

the weight of the plurality of attribute similarity scores is determined by a user.

37. The drug comparator of claim 1,

wherein the one or more name-attribute similarity scores and one or or more product-attribute similarity scores are associated with a weight, and

wherein the drug comparator further comprises an attribute similarity scores processor that applies the weights to their associated attribute similarity scores.

38. The drug comparator of claim 1, wherein, the drug comparator further comprises:

a composite score generator that deterministically generates one or more composite quantitative comparisons based on comparisons of one or more of the attribute-similarity scores.

39. The drug comparator of claim 38,

wherein the one or more composite quantitative comparisons comprise measures of the absolute likelihood of confusing the one or more target drugs with the one or more referencee drugs.

40. A method for assessing the confusability of one or more target drugs and one or more reference drugs based on one or more objective attributes of each target drug and each reference drug, comprising the steps of:

(1) deterministically generating one or more product-attribute similarity scores each representing a similarity of a product attribute of the one or more target drugs and the one or more reference drugs.

41. The method of claim 40, further comprising the step of:

(3) deterministically generating one or more name-attribute similarity scores each representing a similarity of a name attribute of the one or more target drugs and the one or more reference drugs.

42. The method of claim 41,

wherein step (3) comprises the step of

(a) generating one or more name-attribute similarity scores based on one or more orthographic measures of the name attributes of the target and reference drugs.

43. The method of claim 42, wherein the one or more orthographic measures comprises:

one or more N-gram measures.

44. The method of claim 42, wherein the one or more orthographic measures comprises:

one or more edit distance measures.

45. The method of claim 41, wherein:

step (3) comprises the step of;

(a) generating one or more name-attribute similarity scores based on one or more phonetic measures of the name attributes of the one or more target and one or more reference drugs.

46. The method of claim 45, wherein step (3) (a) comprises:

phonetically transforming the name attribute of at least one of the target and reference drugs.

47. The method of claim 46, wherein step (3) (a) further comprises:

applying an orthographic measure to the phonetically reformed name attribute.

48. The method of claim 41, wherein step (3) comprises the step of:

(a) generating one or more name-attribute similarity scores based on one or more phonological measures of the name attributes of the one or more target and reference drugs.

49. The method of claim 41, wherein:

step (3) comprises the steps of

(a) generating one or more name-attribute similarity scores based on one or more orthographic measures of the name attributes of the one or more target and reference drugs;

(b) generating a second one or more name-attribute similarity scores based on one or more phonetic measures of the name attributes of the one or more target and reference drugs; and

(c) generating a third one or more name-attribute similarity scores based on one or more phonological measures of the name attributes of the one or more target and reference drugs.

50. The method of claim 41, wherein:

step (1) comprises any one or more steps selected from the group consisting of

(a) generating at least a first product-attribute similarity score based on one or more dosage strengths of the target and reference drugs;

(b) generating at least a second product-attribute similarity score based on one or more indicated uses of the target and reference drugs;

(c) generating at least a third product-attribute similarity score based on one or more dosage forms of the target and reference drugs;

(d) generating at least a fourth product-attribute similarity score based on one or more administration routes of the target and reference drugs;

(e) generating at least a fifth product-attribute similarity score based on one or more manufacturers of the target and reference drugs;

(f) generating at least a sixth product-attribute similarity score based on one or more pharmacological categories of the target and reference drugs;

(g) generating at least a seventh product-attribute similarity score based on one or more storage requirements of the target and reference drugs;

(h) generating at least an eighth product-attribute similarity score based on one or more colors of the target and reference drugs;

(i) generating at least a ninth product-attribute similarity score based on one or more shapes of the target and reference drugs;

(j) generating at least a tenth product-attribute similarity score based on one or more legal standings of the target and reference drugs; and

(k) generating at least an eleventh product-attribute similarity score based on one or more goods and services trademark descriptions of the target and reference drugs.

51. The method of claim 40, further comprising the step of:

(3) calculating a neighborhood density score for at least a first target drug of the one or more target drugs based on a number of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is a function of the one or more attribute similarity scores.

52. The method of claim 41, further comprising the step of:

(4) mathematically combining a plurality of attribute similarity scores comprising the group of attribute similarity scores consisting of the one or more name-attribute similarity scores and the one or more product-attribute similarity scores to generate a processed attribute similarity score.

53. The method of claim 52,

wherein at least one of the plurality of attribute similarity scores is associated with a weight, and

wherein step (4) comprises applying the weights to their associated attribute similarity scores prior to combining the attribute similarity scores.

54. The method of claim 53, wherein:

wherein the weight of the at least one attribute similarity score is determined by a mathematical/logical function.

55. The method of claim 53,

wherein the step (4) of combining the attribute similarity scores comprises determining a weighted sum of the one or more attribute similarity scores.

56. The method of claim 53, wherein the weight of the plurality of attribute similarity scores is determined by a user.

57. The method of claim 56, wherein the weight of the plurality of attribute similarity scores is determined by a user.

58. The method of claim 52, further comprising the step of:

(5) generating one or more scores for at least one of the one or more target drugs. each severity of confusion score representing the severity of the consequence to the patient of confusing a target drug for a reference drug.

59. The method of claim 58,

wherein, the drug comparator further comprises a composite score generator that deterministically generates one or more severity-weighted similarity scores determined by weighting one or more of the attribute similarity scores by one or more of the severity of confusion scores.

60. The method of claim 59,

wherein the composite score generator further generates a quantitative interpretation that comprises a first set of one or more severity-weighted similarity scores corresponding to a first set of target and reference drugs and a second set of one or more severity-weighted similarity scores corresponding to at least a second set of target and reference drugs.

61. The method of claim 60,

wherein the reference drugs of the first and second sets are the same.

62. The method of claim 59,

wherein the composite score generator generates a quantitative interpretation that compares the processed attribute similarity scores corresponding to a first set of target and reference drugs to a processed attribute similarity scores corresponding to a second set of target and reference drugs.

63. The method of claim 58, wherein:

step (5) comprises generating at least a first of the one or more severity of confusion scores based on a therapeutic index of one or more of the target drugs and a therapeutic index of one or more of the reference drugs.

64. The method of claim 58, wherein:

step (5) comprises generating at least a first of the one or more severity of confusion scores based on a contraindication index of one or more of the target drugs and a contraindication index of one or more of the reference drugs.

65. The method of claim 41,

wherein the one or more name-attribute similarity scores and one or or more product-attribute similarity scores are associated with a weight, and

wherein step (4) comprises applying the weights to their associated attribute similarity scores prior to combining the attribute similarity scores.

66. The method of claim 40, further comprising the step of:

(3) calculating a neighborhood familiarity score for at least a first target drug of the one or more target drugs based on a familiarity of the first target drug and the familiarity of each of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is a function of the one or more name- or product-attribute similarity scores.

67. The method of claim 66, wherein the neighborhood familiarity score is based on a name of the first target drug and names of the first set of reference drugs.

68. The method of claim 67, wherein the first measure of familiarity is determined by using frequency of prescription as a proxy for familiarity.

69. The method of claim 40, wherein the one or more composite quantitative comparisons comprise measures of the absolute likelihood of confusing the one or more target drugs with the one or more reference drugs.

70. The method of claim 40, wherein when a user selects the one or more target drugs, the drug comparator automatically generates the attribute similarity scores.

71. The method of claim 40, further comprising the step of:

(4) mathematically combining a plurality of attribute similarity scores comprising the group of attribute similarity scores consisting of the one or more name-attribute similarity scores and the one or more product-attribute similarity scores to generate a processed attribute similarity score.

72. The method of claim 71,

wherein at least one of the plurality of attribute similarity scores is associated with a weight, and

wherein step (4) comprises applying the weights to their associated attribute similarity scores prior to combining the attribute similarity scores.

73. The method of claim 72, wherein:

wherein the weight of the at least one attribute similarity score is determined by a mathematical/logical function.

74. The method of claim 40,

wherein the one or more name-attribute similarity scores and one or more product-attribute similarity scores are associated with a weight, and

wherein step (4) comprises applying the weights to their associated attribute similarity scores prior to combining the attribute similarity scores.

75. A storage medium that contains software that, when executed on a computing system, performs a method for assessing the confusability of one or more target drugs and one or more reference drugs based on one or more objective attributes of each target drug and each reference drug, comprising the steps of:

(1) deterministically generating one or more product-attribute similarity scores each representing a similarity of a product attribute of the one or more target drugs and the one or more reference drugs;

(2) deterministically generating one or more name-attribute similarity scores each representing a similarity of a name attribute of the one or more target drugs and the one or more reference drugs; and

(3) compositing a plurality of attribute similarity scores comprising the group of attribute similarity scores consisting of the one or more name-attribute similarity scores and the one or more product-attribute similarity scores.

76. The storage medium of claim 75,

wherein step (3) comprises the step of

(a) generating one or more name-attribute similarity scores based on one or more orthographic measures of the name attributes of the one or more target and reference drugs;

(b) generating a second one or more name-attribute similarity scores based on one or more phonetic measures of the name attributes of the one or more target and reference drugs; and

(c) generating a third one or more name-attribute similarity scores based on one or more phonological measures of the name attributes of the one or more target and reference drugs.

77. The storage medium of claim 75, wherein:

step (1) comprises any one or more steps selected from the group consisting of

(a) generating at least a first product-attribute similarity score based on one or more dosage strengths of the target and reference drugs;

(b) generating at least a second product-attribute similarity score based on one or more indicated uses of the target and reference drugs;

(c) generating at least a third product-attribute similarity score based on one or more dosage forms of the target and reference drugs;

(d) generating at least a fourth product-attribute similarity score based on one or more administration routes of the target and reference drugs;

(e) generating at least a fifth product-attribute similarity score based on one or more manufacturers of the target and reference drugs;

(f) generating at least a sixth product-attribute similarity score based on one or more pharmacological categories of the target and reference drugs;

(g) generating at least a seventh product-attribute similarity score based on one or more storage requirements of the target and reference drugs;

(h) generating at least an eighth product-attribute similarity score based on one or more colors of the target and reference drugs;

(i) generating at least a ninth product-attribute similarity score based on one or more shapes of the target and reference drugs;

(j) generating at least a tenth product-attribute similarity score based on one or more legal standings of the target and reference drugs; and

(k) generating at least an eleventh product-attribute similarity score based on one or more goods and services trademark descriptions of the target and reference drugs.

78. The storage medium of claim 75, wherein the method further comprises the step of:

(4) calculating a neighborhood density score for at least a first target drug of the one or more target drugs based on a number of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is a function of the similarity between one or more attribute similarity scores.

79. The storage medium of claim 75, wherein the method further comprises the step of:

(4) calculating a neighborhood familiarity score for at least a first target drug of the one or more target drugs based on a familiarity of the first target drug and the familiarity of each of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is a function of the one or more name- or product-attribute similarity scores.

80. The storage medium of claim 75, wherein:

at least one attribute similarity score is associated with a weight, and the attribute similarity scores processor generates the processed attribute similarity scores at least in part by determining a weighted sum of one or more of the attribute similarity scores.

81. The storage medium of claim 75, wherein the method further comprises the step of:

(4) generating one or more severity of confusion scores representing a severity of confusion between the target and reference drugs.

82. The storage medium of claim 75, wherein when a user selects the one or more target drugs, the drug comparator automatically generates the attribute similarity scores.

83. A computer-implemented system for comparing one or more selected target drugs with one or more selected reference drugs, comprising:

(1) a target-reference data provider that provides target-reference data describing a plurality of attributes of each of the selected target and reference drugs; and

(2) a composite analyzer that, based on the target-reference data, generates one or more composite quantitative comparisons of the selected target and reference drugs, the composite analyzer comprising;

(a) an attribute similarity scores generator that generates one or more attribute similarity scores representing a similarity of attributes of the selected target and reference drugs, wherein the attribute similarity scores generator comprises any one or more attribute comparators selected from the group consisting of

a strength comparator that generates at least a first attribute similarity score based at least in part on one or more dosage strengths of the target and reference drugs;

an indication comparator that generates at least a second attribute similarity score based at least in part on one or more indicated uses of the target and reference drugs;

a dosage form comparator that generates at least a third attribute similarity score based at least in part on one or more dosage forms of the target and reference drugs;

an administration route comparator that generates at least a fourth attribute similarity score based at least in part on one or more administration routes of the target and reference drugs;

a manufacturer comparator that generates at least a fifth attribute similarity score based at least in part on one or more manufacturers of the target and reference drugs;

a pharmacological category comparator that generates at least a sixth attribute similarity score based at least in part on one or more pharmacological categories of the target and reference drugs;

a storage requirements comparator that generates at least a seventh attribute similarity score based at least in part on one or more storage requirements of the target and reference drugs;

a color comparator that generates at least an eighth attribute similarity score based at least in part on one or more colors of the target and reference drugs;

a shape comparator that generates at least a ninth attribute similarity score based at least in part on one or more shapes of the target and reference drugs;

a legal standing comparator that generates at least a tenth attribute similarity score based at least in part on one or more legal standings of the target and reference drugs;

a trademark description comparator that generates at least an eleventh attribute similarity score based at least in part on one or more goods and services trademark descriptions of the target and reference drugs;

or

a name attribute comparator that generates at least a fourteenth attribute similarity score based at least in part on one or more names of the target and reference drugs.

84. The system of claim 83, wherein:

the one or more composite quantitative comparisons are deterministic.

85. The system of claim 83, wherein:

the name attribute comparator comprises any one or more name attribute analyzers selected from the group consisting of

an orthographic analyzer that generates a first of the at least a fourteenth attribute similarity score based at least in part on one or more orthographic measures of the names of the selected target and reference drugs;

a phonetic analyzer that generates a second of the at least a fourteenth attribute similarity score based at least in part on one or more phonetic measures of the names of the target and reference drugs;

or

a phonological analyzer that generates a third of the at least a twelfth attribute similarity score based at least in part on one or more phonological measures of the names of the target and reference drugs.

86. The system of claim 85, wherein:

the composite analyzer further comprises

(b) a composite score generator that generates the one or more composite quantitative comparisons based, at least in part, on one or more of the attribute similarity scores.

87. The system of claim 86, wherein:

the composite analyzer further comprises

(c) a severity of confusion scores generator that generates one or more severity of confusion scores representing a severity of confusion between the target and reference drugs; and

wherein

the composite score generator generates the one or more composite quantitative comparisons based, at least in part, on the severity of confusion scores.

88. The system of claim 86, wherein:

the composite analyzer further comprises

(c) a neighborhood density calculator that calculates a neighborhood density score for at least a first of the selected target drugs based on a number of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is determined by comparing one or more attributes of the corresponding reference drug with corresponding one or more attributes of the first target drug; and

wherein

the composite score generator generates the one or more composite quantitative comparisons based, at least in part, on the neighborhood density score.

89. The system of claim 86, wherein:

the composite analyzer further comprises

(c) a neighborhood familiarity calculator that calculates a neighborhood familiarity score for at least a first target drug of the selected target drugs based on a familiarity of each of a first set of reference drugs, wherein each of the first set of reference drugs has a determined distance from the first target drug that is not greater than a threshold distance, wherein each determined distance is determined by comparing one or more attributes of the corresponding reference drug with corresponding one or more attributes of the first target drug; and

wherein

the composite score generator generates the one or more composite quantitative comparisons based, at least in part, on the neighborhood familiarity score.

90. The system of claim 83, further comprising:

(3) an attribute database updater that updates a drug attribute database comprising data that describes attributes of the target and reference drugs; and

wherein

the target-reference data provider provides the target-reference data based on the updated drug attribute database.