Ethics of Lithotomy

statistics, decision-theory
2014-09-092014-09-09 notes certainty: log importance: 3

Poisson, lithotomy vs lithotrity surgery mortality rates


Civiale back­ground: http://schol­ar.­google.­com/schol­ar?q=­Jean%20­Civiale / http­s://en.wikipedi­­i/­Jean_­Civiale (1836 re­ward & pub­li­ca­tion) Par­al­lèle des divers moyens de traiter les cal­culeux / http­s://­par­al­lledes­di­ver00­civi#­page/348/­mod­e/2up

Acad­e­mie re­port: http://schol­ar.­google.­com/schol­ar?­clus­ter=1883366974866603345&h­l=en&as_s­dt=0,21

where is Civiale’s data? per­haps in Traité pra­tique sur les mal­adies des or­ganes géni­to-uri­naires: Mal­adies du corps de la vessie, Vol­ume 3

1141 of 5715 op­er­ated upon pa­tients died; per­haps can trace full data us­ing these num­bers:

“Com­men­tary: ‘Med­ical art’ ver­sus ‘med­ical sci­ence’: J Civiale’s sta­tis­ti­cal re­search on con­di­tions caused by cal­culi at the Paris Acad­emy of Sci­ences in 1835” Tröh­ler http://i­je.ox­ford­jour­nal­­con­tent/30/6/1252.­full

“Com­men­tary: Treat­ment of blad­der stones and prob­a­bilis­tic rea­son­ing in med­i­cine: an 1835 ac­count and its lessons for the present” http://i­je.ox­ford­jour­nal­­con­tent/30/6/1253.­full Van­den­broucke 2001 “Com­men­tary: The Paris Acad­emy of Sci­ence re­port on Jean Civiale’s sta­tis­ti­cal re­search and the 19th cen­tury back­ground to ev­i­dence-based med­i­cine” http://i­je.ox­ford­jour­nal­­con­tent/30/6/1249.­full Matthews 2001

“Civiale, stones and sta­tis­tics: the dawn of ev­i­dence-based med­i­cine” http://on­lineli­brary.wi­ley.­com/­doi/10.1111/j.1464-410X.2009.08529.x/­full Herr 2009

In the pe­riod from 1823 to 1827, he op­er­ated on 43 pa­tients and claimed 42 suc­cess­es. He called his new method ‘lithotripsy’ and soon be­came con­vinced that it was su­pe­rior to per­ineal litho­tomy [5].

  • 5: Civiale J. De la Lithotri­tie, ou brole­ment de la pierre. Paris, 1827

To prove his point, Civiale be­gan a sta­tis­ti­cal study to com­pare the out­comes be­tween litho­tomy and lithotrip­sy. To do that, he com­piled all op­er­a­tions done by the promi­nent litho­tomists in Eu­rope at the time and tab­u­lated their crude death rates. Lit­tle was known about the re­sults of litho­tomy other than ru­mour, anec­dotes, and prob­a­bly ex­ag­ger­ated good re­sults that amounted to favourable pub­lic­i­ty. Sur­geons re­ceived more pa­tients if they recorded higher suc­cess rates. Civiale did not say how he se­lected the differ­ent cen­tres, how he ap­proached his sur­gi­cal col­leagues, how he col­lected the num­bers and ver­i­fied their ac­cu­ra­cy, who paid him to do so, who had the idea, and why. As to the ‘who and why’, the best guess is that he wanted credit as the in­ven­tor of the first lithotrite and to es­tab­lish that his new blood­less pro­ce­dure for re­mov­ing blad­der stones, lithotrip­sy, was su­pe­rior to the widely used but dan­ger­ous open method of litho­to­my.

…Civiale com­piled a to­tal of 5715 op­er­a­tions for blad­der stone by per­ineal litho­tomy for which the re­sults were well known. Of the­se, he found 1141 deaths (20%) and an­other 100 in­fir­mi­ties be­yond re­pair as a re­sult of the op­er­a­tion. The treat­ment mor­tal­ity rate was one in five, al­though more than half the pa­tients were aged < 14 years and it was known that chances of re­cov­ery at that age were dou­bled. To these re­sults, he com­pared his to­tal of 257 pa­tients treated by lithotrip­sy, of whom there were only six deaths (2.3%), and among these there were only a few aged <14 years. This gave less than one death for 42 pa­tients treated by lithotrip­sy. To com­plete the demon­stra­tion of the su­pe­ri­or­ity of lithotripsy over litho­to­my, he added that since the dis­cov­ery of lithotrip­sy, among a fairly large num­ber of physi­cians suffer­ing from cal­culi, none had re­sorted to litho­to­my; all were op­er­ated by lithotrip­sy. Quite nat­u­rally Civiale keenly prop­a­gated his in­no­v­a­tive method of treat­ment, which he con­sid­ered much safer, as ev­i­denced by stone-re­cov­ery rates of 98% ac­cord­ing to his re­sults, com­pared to 78% he later cal­cu­lated after ag­gre­gat­ing sta­tis­tics on thou­sands of litho­tomies [6].

  • 6: Civiale J. Traite de l’affec­tion cal­culeuse, ou recherches sur la for­ma­tion, les car­ac­teres physiques et chim­iques, les caus­es, les sig­nes, et les effets pathologiques de la pierre et de la grav­elle suives d’un Es­sai de sta­tis­tique sur cette malade. Paris: Crochard et Comp, 1838

How­ev­er, by 1878, it was be­com­ing in­creas­ingly ap­par­ent that lithotripsy had its own dan­gers and mor­tal­i­ty, prob­a­bly be­cause more less-skil­ful sur­geons adopted the method. At that time, litho­tomy still car­ried a mor­tal­ity rate of 20-23%; how­ev­er, the mor­tal­ity rate for lithotripsy now had in­creased from 15 to 23% in pa­tients aged >50 years, and 10% if <50 years. Even the best ev­i­dence does, and should, change.

“Big Data in the 1800s in sur­gi­cal sci­ence: A so­cial his­tory of early large data set de­vel­op­ment in uro­logic surgery in Paris and Glas­gow” http://bd­s.sagepub­.­com/­con­tent/1/2/2053951714543701.­full Mazur 2014

  • Gant F (1878) Sci­ence and Prac­tice of Surgery, 2nd ed. Vol. 2. Philadel­phia, PA: Lind­say & Black­ston, pp. 698-700.

In 1862, the British Med­ical Jour­nal pub­lished a re­port from a hos­pi­tal gazette show­ing that Civiale’s data from 1860 to 1861 fo­cused on both sur­vival/­mor­tal­ity and qual­ity of life. In terms of sur­vival, the fol­low­ing was noted about Civiale’s cas­es:

“Tak­ing all his cases in 1860 and 1861 to­geth­er, M. Civiale has had 120 cal­cu­lous pa­tients: 115 males and 5 fe­males. Of the­se, lithotrity has been per­formed on 88; of whom, 3 have died, 79 have re­cov­ered; and in 6 func­tional dis­tur­bances re­main-in­de­pen­dent, how­ev­er, both of the stone and of the op­er­a­tion. Litho­tomy has been per­formed on 17 pa­tients; of whom 8 are cured, 2 have fis­tu­lae re­main­ing, and 7 have died. The re­main­ing 15 pa­tients have not been op­er­ated on; 6 have died, and 9 are still alive.”

  • Anony­mous (1862). “Progress of med­ical sci­ence”. British Med­ical Jour­nal 1862; 1: 439. Taken from Gazette de Ho­pi­taux, 25 Fevri­er. - not avail­able? http://schol­ar.­google.­com/schol­ar?q=%22Pro­gress+of+med­ical+­science%22+Bri­tish+Med­ical+Jour­nal+1862&bt­nG=&h­l=en&as_s­dt=0%2C21

How­ev­er, not only were Civiale’s own pa­tients in­cluded as data points to sup­port his pro­ce­dure, but his lithotrity pro­ce­dure had also gained wide­spread sup­port from urol­o­gists over Eu­rope who were col­lect­ing data on their pa­tients. Civiale added these sets to his own data in his re­port to the Acad­e­mie des sci­ences. Be­li­naye (1837) notes that with litho­to­my, 1024 of 5443 pa­tients died, and with lithotri­ty, 5 of 245 pa­tients died.

  • Be­li­naye HG (1837) Com­pendium of Lithotripsy, Lon­don: JB Bail­liere, pp. 179

…In terms of Civiale’s pa­tient num­bers, 5443 pa­tients un­der­went surgery for re­moval of the blad­der stone(s), and 245 pa­tients un­der­went Civiale’s lithotrity pro­ce­dure. The Académie des sci­ences com­mis­sioned Pois­son et al. (1835) to eval­u­ate Civiale’s da­ta. Com­men­ta­tors (Black, 2001; Matthews, 2001; Tröh­ler, 2001; Van­den­broucke, 2001) note that this type of pre­sen­ta­tion of data by a sur­geon to an au­gust body of ex­perts - Pois­son, an ex­pert sta­tis­ti­cian; Du­long, a physi­cian-chemist; Lar­rey, a great mil­i­tary sur­geon who was Napoleon’s sur­geon; and Dou­ble, a rep­re­sen­ta­tive of the med­ical es­tab­lish­ment - was a unique form of peer re­view for Civiale’s era. These eval­u­a­tors noted that these num­bers were “all sup­plied by the prac­tice of the great­est sur­geons alive.” Physi­cians in Civiale’s era con­sid­ered the set of 5443+ pa­tient cases a large num­ber. Even though the re­ported num­bers were large, de­rived from ex­perts across Eu­rope, and con­sid­ered even to­day as “hard data” (sur­vival and mor­tal­ity data (Boiv­in, 2014)), the Académie des sci­ences sub­jected Civiale’s data to rig­or­ous re­view.

  • Pois­son SD, Du­long PL, Lar­rey DJ, et al. (1835, 2001) “Sta­tis­ti­cal re­search on con­di­tions caused by cal­culi by Doc­tor Civiale”. In­ter­na­tional Jour­nal of Epi­demi­ol­ogy 30(6): 1246-1249; at p. 1247. Pub­lished in Comtes Ren­dus de l’A­cadémie des Sci­ences, Paris 1835; 1: 167-177. Trans­lated for the In­ter­na­tional Jour­nal of Epi­demi­ol­ogy by An­gela Swaine Verdier.
  • Black N (2001), “Com­men­tary: That was then, this is now”. In­ter­na­tional Jour­nal of Epi­demi­ol­ogy 30(6): 1251.
  • Matthews JR (2001) “Com­men­tary: The Paris Acad­emy of Sci­ence re­port on Jean Civiale’s sta­tis­ti­cal re­search and the 19th cen­tury back­ground to ev­i­dence-based med­i­cine”. In­ter­na­tional Jour­nal of Epi­demi­ol­ogy 30(6): 1249-1250.
  • Tröh­ler U (2001), “Com­men­tary:”Med­ical art" ver­sus “med­ical sci­ence”: J. Civiale’s sta­tis­ti­cal re­search on con­di­tions caused by cal­culi at the Paris Acad­emy of Sci­ences in 1835". In­ter­na­tional Jour­nal of Epi­demi­ol­ogy 30(6): 1252-1253.
  • Van­den­broucke JP (2001), “Com­men­tary: Treat­ment of blad­der stones and prob­a­bilis­tic rea­son­ing in med­i­cine: An 1835 ac­count and its lessons for the present”. In­ter­na­tional Jour­nal of Epi­demi­ol­ogy 30(6): 1253-1258.

Pois­son et al. (1835) noted that these com­par­a­tive cal­cu­la­tions were not made on very ac­cu­rate bases (ret­ro­spec­tive re­view of re­ported cas­es); there­fore, they could not take the place of sci­ence. In ad­di­tion, the ex­pert re­ports lacked de­tails of how the ex­pert urol­o­gists car­ried out the as­sess­ments. Thus, for the Académie des sci­ences, these re­ports fell short of the goal of de­ter­min­ing the nu­mer­i­cal pro­por­tions of mor­tal­ity after in­ci­sion with any de­gree of ac­cu­ra­cy. Yet, the com­mis­sion com­mended Civiale for rec­og­niz­ing both the im­por­tance of large num­bers and many of the prob­lems in­ter­pret­ing large num­bers, es­pe­cially when those large num­bers in­volved a com­pi­la­tion of data sets from differ­ent uro­logic sur­geons. The Académie des sci­ences con­cluded that Civiale’s work com­par­ing sur­gi­cal ap­proaches (surgery ver­sus lithotri­ty) was mov­ing in the right di­rec­tion. In the com­mis­sion’s re­port on Civiale’s data, Civiale was en­cour­aged “to pur­sue his sta­tis­ti­cal re­search to in­crease the vol­ume of data, and to pro­vide more cir­cum­stan­tial de­tail to make it more con­clu­sive…”

As noted by Buchanan, sur­geons of Civiale’s day were al­ready in­ter­ested in both mor­tal­ity data and other data, such as the de­mo­graph­ics of the pa­tient that made a pro­ce­dure like lithotrity more or less suit­able to be un­der­tak­en. From a his­to­ri­an’s per­spec­tive, Tröh­ler (2001) ar­gues that Civiale’s re­port on his lithotrite and lithotrity pro­ce­dure pre­sented to the Académie des sci­ences in 1835 was part of a for­mal eval­u­a­tion of Civiale’s num­bers. This eval­u­a­tion by uro­logic sur­geons and Pois­son, as a par­tic­i­pant eval­u­a­tor, needs to be viewed in “the con­text of a con­test in con­tem­po­rary French med­ical lit­er­a­ture about the ap­plic­a­bil­ity of ‘sta­tis­tics’ to ‘med­i­cine’” (Tröh­ler, 2001). How­ev­er, Civiale’s num­bers on the rel­a­tive mor­tal­ity of his lithotrity pro­ce­dure ver­sus litho­tomy (cys­to­to­my) were so good from a mor­tal­ity re­duc­tion stand­point that ex­pert urol­o­gists through­out Eu­rope adopted the lithotrity pro­ce­dure in their own prac­tice and con­tin­ued with the pro­ce­dure based on their own clin­i­cal ex­pe­ri­ence with both litho­tomy and lithotri­ty. Civiale’s data set on blad­der stone re­moval was the most ex­ten­sive of its day, and his less in­va­sive pro­ce­dure of lithotrity and his in­stru­ment (which other urol­o­gists in France and around Eu­rope used quickly and im­proved upon read­i­ly) con­tin­ued to de­velop over time. Even Civiale started us­ing Heurteloup’s in­stru­ments in his lithotrity prac­tice.