Old Topics‎ > ‎


15-IX-1997 Q1 Decisions on mortality coding

[A point for discussion, from Sue Walker, Australia, with new comments from Canada and England.]

Firstly, can I say how much I am enjoying - and learning from - the mortality discussion group. Thank you Lars Age, for organising it. Although we have not yet implemented ICD-10 in Australia (we are planning to implement ICD-10 for morbidity coding in 1998 and mortality coding in ?1999), I have conducted a number of ICD-10 training sessions in South East Asia and the Western Pacific and the discussions from the group have provided valuable insights for me in understanding the intricacies of mortality coding.

My question is - where we agree to disagree or find problems with the ICD instructions and mortality coding rules, is it planned that we will write a document outlining these concerns to be presented to WHO? or discussed at the Heads of Centres meeting in Copenhagen? or at the next mortality ICE meeting? There seems to be so much valuable experience and information being shared that it would be a pity to lose it. I would recommend that we develop a mortality coding standards manual - much as we have a morbidity coding standards manual in use here in Australia. We publish the standards as volume 4 of our current ICD-9-CM, and have been recently updating the standards for ICD-10-AM (Australian modification). The standards provide guidance to coders where the ICD instructions are not sufficiently clear, where there is not an appropriate index entry or where there is a specific way that we recommend that particular diseases or procedures are coded. This way, even if we are not always correct, at least our coders are being consistent.

I would welcome comments about this suggestion.


*Canada /Patricia Wood/: The idea of an International Nosology Reference Group is just so exciting. It would seem a perfect solution to the issues that Sue Walker raises. I think that all of us participating in this Nosology Forum (ICD-10 e-mail group) would applaud any effort to make some decisions about the things that we discuss. An international mortality coding standard manual, as suggested by Sue, could even be a product of the NRG. I hope that this idea gets resounding support at the Centre Heads meeting in Copenhagen!

I think that classification decisions (i.e. the coding of terms that are not indexed in the ICD-10) are something else that could fall into the realm of responsibility of the NRG. We are all regularly faced with making classification decisions and, I suppose, many of us are making them in isolation. If a list of decisions were to be included in the kind of dynamic standards manual that Sue Walker suggests international consistency would be further enhanced.

*England /Wendy Child/: We, too, have discussed how the Group's deliberations, suggestions and decisions can be presented to WHO and, more importantly, when? The suggestion of a NRG seems to be an excellent one and would be supported here. It is reassuring to know that we are not alone in sometimes having difficulty in interpreting the ICD rules and instructions, and if a consensus of opinion and international agreement can be reached on those points giving concern, and produced somewhere in print, what an achievement!

USA /Julia E Raynor/: I am very excited about the establishment of the Nosology Reference Group. It will make the work of the email group more worthwhile and valuable when the present and future questions can be resolved, not just discussed.

Sweden /Lars Age Johansson/: So far we have just noted that we interpret the ICD differently, and I agree fully with Sue that, if we want to increase international comparability, that is not enough. We also need some mechanism to authoritatively resolve the questions we are discussing. For this year's WHO Centre Heads Meeting (Copenhagen, 14-20 October), Janet Hagey (Statistics Canada) and I are preparing a paper on the discussions in this group. Since it is obvious that there are important differences in the interpretation and application of the ICD, we suggest that the WHO and the Centre Heads establish an international "Nosology Reference Group" (NRG). The NRG would review, and adjudicate, questions on the interpretation of the ICD for mortality coding, and also, when necessary, suggest clarifications of the coding rules. The mortality ICE planning group, through the American WHO Centre, will make a similar suggestion. The discussions in the ICD-10 mortality group would be an important input to the NRG. The deliberations, decisions, and suggestions of the NRG would, of course, be reported to the WHO and the Centre Heads. Hopefully, it would also be possible to produce some kind of mortality coding standards manual, as Sue suggests.

Several other members of our group have talked to me earlier on this, and it would be very interesting to hear your comments.

15-IX-1997 Q2 Meningococcal septicemia due to tonsillitis

(Rule GP, C / UC Chapter I, X)

[A question from Sweden: how to code the following certificate:]

1(a) Acute meningococcal septicaemia
1(b) Tonsillitis

[Suggested coding:]

                  Multiple cause            Underlying cause

*Canada     A392/J039                 A392

England       A392/J039                 A392

*USA         A392/J039                 A392


*Canada /Patricia Wood/: Referring to the notes for Interpretation of "highly improbable" (ICD-10, Volume 2, page 67) I applied Rule 2 to select a tentative underlying cause as A392. In the absence of any linkages or further modifications I selected A392 as the underlying cause of death.

*USA /Julia E Raynor/: We agree with Wendy/England that according to Volume 2, page 67, it is a highly improbable sequence. Although, it seems very likely to me. Septicemia, modified by other organisms are due to many conditions.

England /Wendy Child/: A 'highly improbable' sequence as shown on page 67 Volume 29.

15-IX-1997 Q3 Linkage cerebral infarction - multiinfarction dementia

(Rule C / UC Chapter V, IX)

[Also from Sweden:]

1a Dementia
1b Multiple Cerebral Infarctions

[Suggested coding:]

                         Multiple cause           Underlying cause

Canada             F011/I639                  I639

England             F03/I639                   I639

USA                 F03/I639                   I639


Sweden /Lars Age Johansson/: I made my usual mistake and did not tell you what our problem was - I do not want to influence anybody's coding... But, like Patricia, we were wondering if this is a case a multiinfarction dementia, and, if so, would the exclusion note on p. 496 in Volume 1 apply (I60-I69 ... Excludes ... vascular dementia, F01). Would that give us F011 as the underlying cause of death? I would appreciate your comments on that.

*England /Wendy Child/: My comment was a bit short and probably not very helpful, 'we would accept this sequence'. We did consider making this 'multi infarct dementia' but decided to take into account the note at F01-F09 on page 52 volume 2. I think this means that we will use F01.1 for Underlying Cause when the certificate states the condition only.

Canada /Patricia Wood/: I was debating which code to use for "dementia" as recorded on line a). I was torn between F03 and F011, but either way I selected I639 as the tentative underlying cause of death and, in the absence of any linkages or further linkages I selected I639 as the underlying cause of death.

USA /Julia E Raynor/: General principle; no linkage

England /Wendy/: We would accept this sequence.

15-IX-1997 Q4 Multiple valvular disorders due to arteriosclerosis

(Rule C / UC Chapter IX)

[And again a question from Sweden:]

1a Mitral and Aortic Stenosis
1b Generalized arteriosclerosis

[Suggested coding:]

                     Multiple cause           Underlying cause

*Brazil           I342 I350/I709         I342

Canada          I342 I350/I709         I342

England          I342/I350/I709         I342

USA             I342 I350/I709           I342


*Brazil /Augusto Hasiak Santo/: Rules: General Principle and Rule C (after conflict resolution).

Canada /Patricia Wood/: I am terribly embarrassed to send an "update" to my original response to this question! Clearly, when coding mitral and aortic stenosis specified as due to generalized arteriosclerosis, ICD-10 codes I342 and I350 are the correct choices. I would apply the General Principle to select I709 as the tentative underlying cause of death. I709 links with both I342 and I350 (Volume 2, page 57) creating a conflict in linkage which would be resolved by linking with I342, the condition that would have been selected if the cause initially selected had not been reported (Volume 2, page 46). In the absence of any further modifications or linkages I would select I342 as the underlying cause of death.

Canada /Patricia Wood/: This is a tough one. Stenosis, mitral (valve), with aortic valve disease, is indexed to I080. I080 includes disorders of both mitral and aortic valves whether specified as rheumatic or not. Referring to the notes for Interpretation of "highly improbable" (ICD-10, Volume 2, page68) it is stated that rheumatic heart disease (I05-I09) can not be due to "any disease other than ..."

I applied Rule 2 to select a tentative underlying cause of I080 and, in the absence of any linkages or further modifications, I selected I080 as the underlying cause of death.

USA /Julia E Raynor/: Select generalized arteriosclerosis by General Principle. Next, according to linkage in Volume 2, page 57, arteriosclerosis links in a "due to" position (LDP) with mitral stenosis, I342 and aortic stenosis, I350. The classification has not provided a combined code for both valves when non-rheumatic nor established a priority order of one valve over another for ICD 10. Therefore, there is a conflict in linkage. We would follow our general instruction of linking with the condition we would have chosen as uc if the selected uc had not been reported. This instruction in included in NCHS Instruction Manual, Part 2a, page 31.

England /Wendy Child/: Using the General Principle, we would select Generalized arteriosclerosis, then modification Rule C to give I342. There is no combination code for mention of both valves due to a non-rheumatic cause and no note to suggest order of priority as in ICD-9, therefore we would select the first mentioned.

6-X-1997 Q1 – 9

6-X-1997 Q1 Coronary atherosclerosis due to cor pulmonale (Rule GP / UC Chapter IX)

6-X-1997 Q2 Angina due to bronchitis  (Rule GP / UC Chapter IX, X)

6-X-1997 Q3 Urinary infection obvious consequence of multiple sclerosis  (Rule 3 / UC Chapter VI, XIV)

6-X-1997 Q4 Viraemia due to pneumonia  (Rule GP / UC Chapter I)

6-X-1997 Q5 Intestinal obstruction obvious consequence of Crohn's disease  (Rule 3 / UC Chapter XI)

6-X-1997 Q6 Staphylococcal infection due to meningitis (Rule GP / UC Chapter I)

6-X-1997 Q7 Chronic glomerulonephritis obvious consequence of hypernephroma  (Rule 3 / UC Chapter II, XIV)

6-X-1997 Q8 Cardiorenal failure due to hypertension NOS  (Rule C / UC Chapter IX)

6-X-1997 Q9 Anaemia due to menorrhagia  (Rule B / UC Chapter III, XIV)

[Further comments from Canada:]

[Suggested coding:]

                        Multiple causes                                Underlying cause

Brazil               I251/I279/J42                                 I251

*Canada         I251/I279/J42                                  I251

England                                                                   I251

Norway                                                                   I251

Sweden                                                                  I251

USA            I251/I279/J42                                     I251

Brazil           I509/I209/J40 J439                            I209

*Canada       I509/I209/J40 J439                           I209

England                                                                 I209

Norway                                                                 J448

Sweden                                                                  J448

USA             I509/I209/J40 J439                           I209

Brazil            N19/N19/N390*G35                        N390

*Canada       N19/N19/N390*G35                       N390

England                                                                 G35

Norway                                                                 G35

Sweden                                                                 N390

USA              N19/N19/N390*G35                     N390

*Canada        B349/J189                                      B349

Brazil             B349/J189                                      B349

England                                                                 B349

Norway                                                                J129

Sweden                                                                B349

USA             B349/J189                                       B349

Brazil           J690/K566*K509                            K509

*Canada        J690/K566*K509                          K566

England                                                                K509

Norway                                                               K509

Sweden                                                               K509

USA             J690/K566*K509                          K566

Brazil            A490/G039                                    A490

*Canada       A490/G039                                   A409

England                                                              A490

Norway                                                             G003

Sweden                                                              A490

USA            A490/G039                                   A490

Brazil           R402/N189/N039*C64                N039

*Canada     R402/N189/N039*C64                  N039

England                                                              C64

Norway                                                              C64

Sweden                                                             N039

USA          R402/N189/N039*C64                  N039

Brazil         I509/I10/I709                                  I110

*Canada     I509/I10/I709                                I132

England                                                             I132

Norway                                                            I132

Sweden                                                            I132

USA           I509/I10/I709                                I132

Brazil          D649/N920                                   N920

*Canada     D649/N920                                  N920

England                                                            D500

Norway                                                           D649

Sweden                                                          N920

USA         D649/N920                                   N920


1(a) Coronary atheroma
1(b) Cor pulmonale
1(c) Chronic bronchitis

*Canada /Patricia Wood/: Although not specifically mentioned in the Interpretation of "highly improbable" (Volume 2, page 67) I would consider coronary atheroma due to cor pulmonale or chronic bronchitis to be unacceptable causal relationships (based on ICD-9 ACME Decision Tables) and I would apply Rule 2 to select I251 as the tentative underlying cause of death. In the absence of any further modifications or linkages I would select I251 as the underlying cause of death.

Brazil /Augusto Hasiak Santo/: Rule 2. Neither cor pulmonale nor chronic bronchitis can be accepted as leading to coronary atheroma.

England /Wendy Child/: Rule 2: we would not accept an atheroma due to cor pulmonale or respiratory diseases.

USA /Julia E Raynor/: Rule 2. We would not accept an atheroma or an atherosclerotic condition due to cor pulmonale or chronic bronchitis.

Norway /Anne Gro Juelsen/: No sequence. Rule 2. (Coronary atheroma is not due to Cor pulmonale.

Sweden /Lars Age Johansson/: I cannot see how a pulmonary disease could cause an atheroma, so, like my Norwegian colleagues, I would use Rule 2 and select the coronary atheroma.


1(a) Acute cardiac failure
1(b) Angina
1(c) Bronchitis and emphysema

*Canada /Patricia Wood/: Although not specifically mentioned in the Interpretation of "highly improbable" (Volume 2, page 67) I would consider angina due to bronchitis or emphysema to be unacceptable causal relationships (based on ICD-9 ACME Decision Tables) and I would apply Rule 1 to select I209 as the tentative underlying cause of death and in the absence of any further modifications or linkages I would select I209 as the underlying cause of death.

Brazil /Augusto/: Rule 1. Bronchitis and emphysema would produce an obstructive pulmonary disease and the heart disease would be due to cor pulmonale. Angina is produced directly by coronary disease.

[This is no doubt the use of the term "angina pectoris" we would like to see, but might it not happen that it is used of chest pain from other causes, like the mechanism Anne Gro describes? This leads to a general question: should we code strictly according to the "correct" usage of the terms, or should we sometimes accept that terms might have been used in a somewhat broader, and perhaps not quite correct, sense? /Lars Age]

England /Wendy/: Rule 1: we do not accept angina due to respiratory conditions but the reason for this is lost 'in the mists of time', I'm afraid!

Norway /Anne Gro/: We have seen the answers from Norway and Sweden and do not quite understand #2, bronchitis and emphysema. Lars Age has selected J40 for the underlying cause, while we had J448, following the instruction on p. 65 in Volume 2 (J40 with mention of J43 links to J 44). Should not the Notes always be applied?

Sweden /Lars Age/: This is very embarrassing. Of course the Notes in Volume 2 should always be applied. The only excuse I can offer is that I am now in that dangerous phase when you start to think that you know the Notes so well that you do not have to look everything up. As any ICD teacher would tell you, you should never think that you know anything in the ICD by heart! I have at least given you a good example of that...

USA /Julia /: Rule 1. We would not accept angina, I209, as due to respiratory diseases probably because it is not included under terminal circulatory disease in Volume 2, pages 68 and 69.

Norway /Anne Gron/: We think we will accept this as a sequence. U.C. J44.8 (bronchitis and emphysema linked together). Angina (we read angina pectoris) is a symptom of insufficient oxygen to the heart muscle and can be caused by the lung disease (the heart must work harder, it grows bigger and...) and in the end we have the cardiac failure.

Sweden /Lars Age/: For ICD-9, we accepted cardiac ischaemia as due to a number of diseases that would diminish the flow of oxygen to the heart. Anne Gro sounds convincing, so until someone tells me that angina cannot be caused by bronchitis with emphysema, I would accept her line of argument and use Rule 1 here. My underlying cause code is J40 (Bronchitis, not specified as acute or chronic) rather than J44 (Other chronic obstructive pulmonary disease, resulting from combining *chronic* bronchitis with emphysema). As far as I can see, it is not necessary to presume that the bronchitis was chronic, since an acute cardiac failure might be caused by an acute myocardial ischaemia.


1(a) Uraemia
1(b) Renal failure
1(c) Urinary tract infection
11 Multiple sclerosis

*Canada /Patricia/: I would apply the General Principle to select N390 as the tentative underlying cause of death. In the absence of any further modifications or linkages I would select N390 as the underlying cause of death. I agree that there is a possibility of a direct consequence relationship with the urinary tract infection and the multiple sclerosis but I would not assume it.

Brazil /Augusto/: General Principle. It seems that there is not a clear evidence to apply Rule 3 and consider urinary tract infection as a direct consequence of multiple sclerosis.

England /Wendy/: Rule 3: we see the UTI as a direct consequence of MS probably due to a catheter?

USA /Julia /: General principle.

Norway /Anne Gro /: According to Norwegian coding practice we would see the urinary tract infection here as a direct consequence of MS and use Rule 3. U.C G35, but in case we would ask back to the doctor regarding duration of MS and kidney disease (our new system makes it very easy to ask back to the doctor).

Sweden /Lars Age/: We would use only the General Principle in this case, and not apply Rule 3. The reason is that even if multiple sclerosis often is a serious condition, that is not true for all cases of the disease. We would, therefore, not use Rule 3 unless there is some evidence of disabling complications. Like the Norwegians, we would ask for more information, but we would stick to Part I if we could not get any answer from the certifier.


1(a) Viraemia
1(b) Pneumonia

*Canada /Patricia / : As specified in the Interpretation of "highly improbable" (Volume 2, page 67, (a)) I would consider viremia due to pneumonia to be an unacceptable causal relationship and I would apply Rule 2 to select B349 as the tentative underlying cause of death. In the absence of any further modifications or linkages I would select B349 as the underlying cause of death. I would not link the viremia with the pneumonia (to yield viral pneumonia) because I think that would imply a causal relationship that has not been certified (i.e. pneumonia due to viremia).

Brazil /Augusto/: Rule 2. This is a death certificate in which the underlying cause has changed between the Ninth and Tenth Revisions. Notice that viraemia was coded as 790.8, an ill-defined code in the Ninth Revision, and was accepted as due to pneumonia, coded as 486, then selected as the underlying cause. Rule D would be applied if pneumonia had been mentioned due to viraemia. This is a problem to be discussed and clarified. Should viraemia, bacteremia and fungemia receive the same interpretation as septicaemia and considered as an exception in the note regarding infectious and parasitic diseases as "highly improbable" ?

England /Wendy/: Rule 2: but hotly debated here. Our previous coding practice allowed us to use specificity, but as viraemia is now in the infectious disease chapter we agreed (?!) it is no longer General rule so specificity cannot be applied (is there a case for viraemia to be treated same as septicaemia?)

USA /Julia /: Rule 2. This is not a sequence per "highly improbable" list in Volume 2, page 67.

Norway /Anne Gro /: Volume 2, 4.2.2, a highly improbable- gives us Rule 2 and Viremia, then we use rule D (specificity) and end up with U.C J12.9.

Sweden /Lars Age/: The sequence is difficult to accept as the "Highly improbable" section now stands. I think I would keep B349 as the underlying cause. Linking it with pneumonia to viral pneumonia would imply that the sequence is after all acceptable, which the "Highly improbable" section rules out.


1(a) Inhalation pneumonia
1(b) Intestinal obstruction
c -
11 Crohns disease

*Canada /Patricia /: I would apply the General Principle to select K566 as the tentative underlying cause of death. In the absence of any further modifications or linkages I would select K566 as the underlying cause of death. As with Q3 I agree that there is a possibility of a direct consequence relationship in this case with the intestinal obstruction and the Crohn's disease but I would not assume it.

Brazil /Augusto /: General Principle + Rule 3. Automatic coding provided intestinal obstruction (K56.6) as the underlying cause. Nevertheless, pathology books include intestinal obstruction as a complication in the natural history of Crohn's disease. Consequently the codes for intestinal obstruction will be considered as direct consequence of Chron's disease.

England /Wendy/: Rule 3: a most definite direct sequel as far as all our medical advisors are concerned.

USA /Julia /: General principle.

Norway /AnneGro/: Rule 3. U.C is K50.9

Sweden /Lars Age/: I agree.


1(a) Staphylococcal infection
1(b ) Meningitis

Brazil /Augusto/: Rule 2

*Canada /Patricia /: As specified in the Interpretation of "highly improbable" (Volume 2, page 67, (a)) I would consider staphylococcal infection due to meningitis to be an unacceptable causal relationship and I would apply Rule 2 to select A490 as the tentative underlying cause of death. In the absence of any further modifications or linkages I would select A490 as the underlying cause of death. As with Q4 I would not link the staphylococcal infection with the meningitis (to yield staphylococcal meningitis) because I think that would imply a causal relationship that has not been certified (i.e. meningitis due to staphylococcal infection).

England /Wendy/: Rule 2: Again opinion was divided as to whether we can use specificity, debate continues!

USA /Julia /: Rule 2. This is not a sequence per "highly improbable" list in Volume 2, page 67.

Norway /Anne Gro /: Same as 4. Rule 2 and rule D. U.C is G00.3

Sweden /Lars Age/: As in no. 4, I agree with Anne Gro that we will have to cut the sequence and use Rule 2 to select the staphylococcal infection as the underlying cause. And again, I find that I cannot combine the two conditions using Rule D since I have rejected the sequence. That would give A49.0 as the underlying cause of death, which I am not too happy with - it seems rather obvious that the certifier had a staphylococcal meningitis in mind. On the other hand, one might argue that the infection on line I(a) might be an infected skin ulcer due to immobility caused by the meningitis. And if we do not know the exact relationship between the conditions, we have better not to presume anything and to code exactly as written. Section 4.2.2. (a) - infections as "highly improbable" consequences of other conditions - says that in principle, an infectious disease may not be caused by a condition outside Chapter I. It then goes on to list a number of exceptions. I wonder if we would not need more exceptions, e.g. infections often seen as complications to other diseases?


1(a) Coma
1(b) Chronic renal failure
1(b) Chronic glomerulonephritis
11 Hypernephroma

*Canada /Patricia /: I would apply the General Principle to select N039 as the tentative underlying cause of death. In the absence of any further modifications or linkages I would select N039 as the underlying cause of death.

Brazil /Augusto /: General Principle

England /Wendy/: Rule 3: Lars' comments made us think, but we still think this is a Rule 3 situation.

USA /Julia /: General principle.

Norway /Anne Gro /: Rule 3. U.C is C64, but like question 3 we would ask about the duration of the diseases.

Sweden /Lars Age/: I would use the General Principle, but not Rule 3, since I do not really see how a *chronic* nephritis could be caused by a hypernephroma. Is it not more likely that the chronic nephritis existed before the malignant neoplasm?


1(a) Cardiorenal failure
1(b) Hypertension
1(c) Arteriosclerosis

*Canada /Patricia /: I would apply the General Principle to select I709 as the tentative underlying cause of death. Then I would apply Rule C to reselect I10 (Volume 2, page 57) as the tentative underlying cause of death.

Next I would apply Rule C to reselect I110 (Volume 2, page 54) as the tentative underlying cause of death. In the absence of any further modifications or linkages (in the mortality classification notes in Volume 2) I would select I110 as the underlying cause of death. BUT, I agree that the correct code for hypertensive cardiorenal failure is I132 and that this should in fact be the underlying cause of death for this record. This is, in my opinion, a point requiring "clarification."

Brazil /Augusto /: Rules GP+RC+RC. Using the Tenth Revision Index and automatic coding, the underlying cause appears as I11.0 (hypertensive heart disease with heart failure) and the renal failure is not taken into account. Notice that cardiorenal failure in the Ninth Revision was coded as 404.9 (hypertensive heart and renal disease) and the final underlying cause for the same death certificate would correspond to I13.2 in the Tenth Revision. The question that arises is if I50.9 correctly represent cardiorenal failure for which the answer implied in the resolution of this death certificate is not affirmative. The confirmation of the code I50.9 for cardiorenal failure will demand, for automatic processing, some kind of parallel procedures for the correct identification of the underlying cause, such as created artificial codes, reject messages or decision through dialogue boxes. The intensive use of these measures that leads to manual coding neutralize and nullify the advantages of automatic programs

England /Wendy/: General Principle and Modification C: The final link to I132 is not easy to find. As Lars says it is important to look at the meaning of the words not just the 'numbers'.

USA /Julia /: Select arteriosclerosis, I709, by general principle. In Volume 2, page 57, arteriosclerosis, I709, links with mention of hypertension, I10. In Volume 3, cardiorenal failure, hypertensive, I132.

Norway /Anne Gro /: I70.9 with mention of I10 gives us I10. Volum 3 - Index: Failure, cardiorenal, hypertensive will give us U.C = I13.2 But we don't quite understand that Failure, cardiorenal (chronic) is coded to I50.9. If we use volum 2 p. 62 table 1 and link I10 as cause of  I50.- the resulting linked code is I11.0. Can anybody explain that?

Sweden /Lars Age/: Using the General Principle and the linkages Anne Gro specifies I will also arrive at I13.2 for the underlying cause. I suppose that the apparent conflict between I13.2 for "cardiorenal hypertensive failure" and I11.0 resulting from the code for "hypertension" linked to the code for "cardiorenal failure", shows that just using the codes for the multiple conditions will not always produce the correct linked code for the underlying cause of death - you must also check what the codes stand for, and make sure that some element which might effect the coding does not get lost on the way. (I hope I do not sound too schoolmasterly, it took me some time to discover this myself.)


1(a) Anaemia
1(b) Menorrhagia

*Canada /Patricia /: Rule B/Rule 6 ... my pet peeve! I would apply the General Principle to select N920 as the tentative underlying cause of death. I would consider Rule B since the tentative underlying cause of death is a trivial condition unlikely to cause death BUT I would decline to apply Rule B because the trivial cause is reported as the cause of more serious complication and therefore not so trivial after all! In making this decision I am however making the very debatable practice of using the wording to the ICD-9 Rule 6b) and I am still wondering if the rewording of this rule in ICD-10 was intended to change its application.

Brazil /Augusto /: General Principle. Rule B does not apply since the trivial affection has given rise to another severe one.

England /Wendy/: General and Modification D: We specified the underlying cause as haemorrhagic anaemia.

USA /Julia /: General principle.

Norway /Anne Gro /: We are not sure here, but we are tempted to use Rule B, trivial condition (Menorrhagia is not likely to cause death ??), reselect and we have U.C D64.9 anaemia. We would definitely ask back to the doctor in this case.

Sweden /Lars Age/: We had a discussion on Rule B some months ago. Since then, I find that I have changed my mind, and now think that since the death was actually caused by a complication to a normally, but not always, harmless condition, then that condition is by definition not trivial. There is an acceptable sequence on the certificate, so we cannot say that the menorrhagia is unlikely to cause death - it did so in the case we are coding. But apparently we need some clarification here...

6-X-1997 Q10 "And" and "with" in titles

(UC Chapter XIX)

[An answer from Canada:]

[Wendy's question was:]

We are also having trouble with Chapter XIX , in particular interpreting the note on Page 891 Volume 1, "where multiple sites of injury are specified in the titles, the word "with" indicates involvement of both sites, and the word "and" indicates involvement of either or both sites". Can anyone give us a simple explanation?! (We did not find TENDON very helpful on this subject).


*Canada /Patricia Wood/: I agree with Lars Age that the note on page 891 refers to code category and subcategory titles that specify multiple sites of injuries like S00.5 Superficial injury of lip 'and' oral cavity and T00.0 Superficial injuries involving head 'with' neck. Referring to the note on page 891 I infer the former code (S00.5) to include superficial injury of the lip only, superficial injury of the oral cavity only, and superficial injury of both the lip and oral cavity together (i.e. lip and/or oral cavity). On the other hand, I infer the latter code (T00.0) to include only superficial injuries of both head and neck together (i.e. neither head nor neck only).

USA /Julia E Raynor/: I am very confused also, but will try to explain our interpretation of note on page 891 and give some examples. When injury (of a site) or specified type of injury (of a site) is stated multiple, bilateral, both, code as indexed under multiple. We do not combine and code multiple injuries of more than one site to a single code. We will not consider the plural form of injuries as "multiple".

1. Example: multiple crushing injuries to both thighs T043

Code Injuries, multiple, limb, lower T043

2. Example: crushing injuries, both thighs T043

3. Example: crushing injuries thighs S771

4. Example: multiple dislocations of wrist, elbow and neck

T032 T032 S133

5. Example: dislocations of wrist, elbow and neck S630 S531


Sweden /Lars Age Johansson/: I think this means that "and" in a rubric stands for "and/or", so that eg. S39.0, "Injury of muscle and tendon of abdomen, lower back and pelvis" covers injuries to muscle(s) of abdomen, injuries to muscle(s) of lower back, injuries to muscle(s) of pelvis, injuries to muscle(s) of both abdomen and lower back, injuries to muscle(s) of both abdomen and pelvis, and injuries to muscle(s) of all three sites. In the same way, the code would cover injuries of tendon(s) to either one, two or three of these sites, with or without injuries to the muscle(s). On the other hand, if the rubric uses "with", then you should not use the code unless all the sites and injuries mentioned in the rubric are present, e.g. T02.1, "fractures involving thorax with lower back and pelvis", should be used only if the certificate mentions both thorax and lower back.

Perhaps this was not a very simple explanation...

6-X-1997 Q11 Coding of multiple injuries

(UC Chapter XIX)

[Here is a comment from Patricia Wood:]

[The question, again from Wendy Child, was:]

1(a) Fracture skull and internal injuries Driver of motor vehicle in collision with another motor vehicle

Our problem is coding a combination of different TYPES of injuries to different body regions, in the absence of a 'order of preference' list as in ICD-9.

[To make our discussions a bit more complicated, I have made up a similar certificate: /Lars Age/]

1(a) Fracture skull and internal injuries of thorax

Driver of motor vehicle in collision with another motor vehicle

[Suggested coding:]

                                Multiple causes                       Underlying cause

*Canada  (a)            S029 T148*V435                   V435
               (b)            S029 S279*V435                   V435

Sweden   (a)            S0290 T148 &V435               V435 S0290
               (b)           S0290 S27990 &V435            V435 T068

USA       (a)            S029 T148*V877                    V877


*Canada /Patricia Wood/: I would code each of the two injuries independently and as indexed arriving at codes S029 and T148. I understand Wendy's concern for no "order of preference" list. In Canada, we include a Nature of Injury table in our annual mortality statistics so we select one nature of injury (N code) for each external (E code) underlying cause of death. We make frequent use of the order of preference list in ICD-9 to select a nature of injury when more than one has been certified without a clear indication as to which one caused death.

For the second scenario, I would code each of the two injuries independently and as indexed arriving at codes S029 and S279. Again, I would need to select one of the nature of injury codes for tabulation. I am not completely sure how we are going to handle this situation in the absence of an "order of preference" list ... that's a problem to solve another day!

Looking at the external cause code I selected V435 assuming the motor vehicle to be an automobile or car. Reviewing Julia's comments about her choice of external cause code I am really concerned. I agree with her interpretation of the ICD-10 transport categories but I think that in Canada "motor vehicle" is widely used synonymously with "car". We (in Canada) will definitely have to figure this out, particularly since we are intending to use the MICAR/ACME system for ICD-10.

USA /Julia E Raynor/: The code assignment of V877 for driver of motor vehicle in collision with another motor vehicle is based on our interpretation of the 10th revision transport categories. Since the main axis of land transports is the victim's mode of transport, it is necessary to determine which category the reported vehicle belongs in. Definitions related to transport accidents, Volume 1, pages 1018-1023 define some types of transports, (car, pick-up etc.), but does not specifically address motor vehicle, so stated. We have not found any instructions to indicate that a motor vehicle is synonymous with car or automobile. We have found index entries that indicate motor vehicle is not synonymous with car. For example, Volume 3, page 579, Accident, transport, occupant, vehicle, motor (traffic) NEC V892 compared with Accident, transport, occupant, automobile V499. We interpret this to mean that motor vehicle, so stated, is not synonymous with car. Another example of reference to motor vehicle is Accident , transport, person, collision, motor vehicle V877. We have not found provision in Volume 3 for coding specific types of accidents involving "motor vehicles". Perhaps we have overlooked some information and would appreciate other opinions.

*Sweden /Lars Age Johansson/: It does not seem to me that Julia has overlooked anything. My reason for coding the "motor vehicle" as a car is that Swedish forensic pathologists use the corresponding expression only of cars. Perhaps that is not true for all countries.

USA /Julia /: In the U.S., we would code the injuries separately by type and site and not consider order of preference.

Sweden /Lars Age/: I am glad that Wendy brought this up, because we in Sweden have also found the multiple injuries very difficult to code.

First, there seems to be a contradiction between the Note on p. 308 in Volume 3 (the second column, at "injury - multiple sites"), which says that multiple injuries classifiable to the same rubric or block should be coded to the appropriate subcategory for multiple injuries within that rubric or block, and the instructions and examples on pp. 86-87 in Volume 2, which tells us to apply the selection rules as usual, if a single body region is mentioned on the certificate. After same hesitation - and much confusion - we at Statistics Sweden decided to follow Volume 2. We would much like to have your comments on that.

In the first example (the certificate Wendy sent us) I would say that "internal injuries" is ill-defined - it might describe injuries to the brain. In that case I would apply Rule D, specificity, and select the skull fracture. If, as in the second certificate, it is quite clear that the internal injuries were not in the head, I cannot see how we could end up with anything else than T06.8 (other specified injuries involving multiple body regions), which is rather disappointing since it does not give much information on the nature of the injury. But the instruction on p. 87 tells us to use T00-T06 if more than one body region is involved, and I cannot find any "better" code than T06.8.

6-X-1997 Q12 Chronic obstructive airway disease NOS plus asthma/emphysema

(UC Chapter X)

[Comments from Canada - perhaps we can at least agree that there is some need for clarification from the WHO here?/lj]

Do you agree that there is an inconsistency between the Index and Volume 1 for obstructive chronic airway disease NOS mentioned together with some more specified form of chronic airway disease, like asthma or emphysema? For example: In the alphabetical index, the code for "Disease - lung - obstructive - with asthma" is J44.-, but in Volume 1, the exclusion note for J44 says that asthma is excluded from J44. We have the same problem with emphysema and some forms of chronic bronchitis. So how would you code these two certificates:

1(a) Obstructive airway disease (asthma bronchiale)
1(a) Pneumonia
1(b) Chronic obstructive airway disease with emphysema

[Suggested coding:]

                 Multiple causes          Underlying cause
USA                                           J448
USA        J189/J449 J439           J448


*Canada /Patricia Wood/: I am not submitting codes for the scenarios presented in this question because I am so thoroughly confused that I can barely tell up from down, never mind J44 from J45!!!

To begin, I can not seem to get beyond problems in the index to address the problems in the questions! I notice that in the index "chronic" is treated as an optional modifier in the diagnostic statement "chronic obstructive lung disease" but not for "chronic obstructive airway disease" so right away

I have a difficulty coding "obstructive airway disease" when not specified as "chronic." Further to that, "chronic obstructive airway disease" is not cross-referenced to "(chronic) obstructive lung disease" so I wonder if all the "with asthma ... with emphysema ..." indexing is applicable. Am I just being far too picky (or perhaps ... literal!?!)?

Now indexing problems aside ... if Q12a were certified as 1(a) obstructive lung disease (asthma bronchiale) I would still have a dilemma. I am inclined to code multiple causes to J449 J459, then to apply Rule 2 to select J449 as the tentative underlying cause and then Rule D to reselect J459 as the underlying cause considering "asthma bronchiale" as a more specific description about the nature of the "obstructive lung disease." However, the exclusion note at J45 (excluding chronic obstructive asthma) and the comparable inclusion note at J44 make me question this choice.

Lars Age Johansson presents a similar scenario in A: 6-X-1997 Q12, update 1 with the statement "chronic obstructive lung disease (emphysema)." As above, I am inclined to consider the multiple causes to be J449 J439 with Rule 2 and Rule D applying to select J439 as the underlying cause of death. Although I do not have the inclusion/exclusion note problem, doubt is cast on this decision if the indexing "chronic obstructive lung disease with emphysema" is considered in which case the code is J44-.

Finally, for Q12b, if certified as 1(a) obstructive lung disease with emphysema, there can not be much doubt that the index indicates "chronic obstructive lung disease with emphysema" should be coded to J44-. However the exclusion note at J44, excluding emphysema, gives me pause.

There, how's that for a completely unhelpful contribution?!?

Sweden /Lars Age Johansson/: Julia is obviously right on the asthma, the inclusion note on chronic obstructive asthma (Vol. 1, p. 535) makes that clear. What I still find confusing is that the statement "Chronic obstructive lung disease (emphysema)" - an ill-defined term followed by a more precise one - according to the Index entry on page 167 goes to J44.-, but that "chronic obstructive emphysema", according to the entry on page 203, goes to J43.9. To me, these two statements are equivalent. Perhaps we should try to get an indication from the WHO if that is what they intended, and if so, why (I can envisage some difficulties in explaining that coding to researchers).

USA /Julia E Raynor/: 12 a - I don't interpret the index and Volume 1 as an inconsistency. I understand the linkage in the index as meaning that (chronic) obstructive lung disease, J449, links with mention of with combination code (LMC) asthma and the combined code is J448, other specified chronic obstructive pulmonary disease. I interpret the exclusion note under category J44 as meaning the code assignment for asthma without chronic obstructive lung disease is J45. Note that J44 includes chronic obstructive asthma. I would treat the linkage with chronic obstructive airway disease the same as chronic obstructive lung disease even though not indexed under chronic obstructive airway disease.

12b - Select chronic obstructive airway disease, J449 and link with mention of combination code emphysema for combined code of J448.

6-X-1997 Q13 Liver cirrhosis with mention of alcohol

(UC Chapter XI)

[A comment from Canada:]

In ICD-9, the rubric of 571.5 was "cirrhosis of liver without mention of alcohol", so we coded all cirrhoses to 571.2 (alcoholic cirrhosis of liver) if alcohol was mentioned on the certificate, either in Part 1 or Part 2. In ICD-10 the wording is changed, and the non-alcoholic category, K74.6, now has the rubric "Other and unspecified cirrhosis of liver". I have not been able to find anything in either the Index or in Volume 2 that links cirrhosis of liver with a mere mention of alcohol. So, how would you code:

1(a) Hepatorenal syndrome
1(b) Liver insufficiency
1(c) Cirrhosis of liver
2 Ethylism

[Suggested coding:]

                    Multiple causes                       Underlying cause

Brazil           K767/K729/K746*F102        K703

*Canada       K767/K729/K746*F102       K703

USA             K767/K729/K746*F102       K703


*Canada /Patricia Wood/: I agree with Julia that, in spite of the change in the ICD-10 subcategory title, we should continue to link liver cirrhosis (K746) with any mention of alcohol use (F10) to yield alcoholic liver cirrhosis (K703). I think that this would be an application of Rule C considering the index to make provision for such a linkage.

Brazil /Augusto Hasiak Santo/: Rules: General Principle + Rule 3 + Rule C (or GP + DSC). Cirrhosis of liver can be considered as a direct consequence of ethylism and as such the entry on the Index for "cirrhosis (hepatic), alcoholic" might be applied.

USA /Julia E Raynor/: The U.S. will link cirrhosis, K746, with mention of alcohol for combination code of K703 since indexed alcoholic. We have decided to make this linkage because of the close relationship between cirrhosis and alcohol even though title of category is different in 10th than 9th.