PUBBLICAZIONI

1) A case of Hellp syndrome with a dramatic course successfully treated with early plasmapheresis.

2) Carbamazepine poisoning treated with plas-mapheresis. Case report.

3) ACUBASE a new clinical tool for quality of care evaluation in Intensive Care

4) Percutaneus dilatational tracheostomy (PDT) and difficult airways: is fibrotracheoscopy (FTS) always able to avoid complications ?

5) A comparison between a dynamic ( OFS) and a static (SAPS II) severity disease index

6) Comparison of complications between percutaneous and translaryngeal tracheostomy: a 3 years experience.

7) SIRS in ICU: a different approach in sepsis assessment: clinical predictivity, severity scores, costs

8) Valutazione clinica del dosaggio degli antibiotici aminoglicosidici e glicopeptidici in Terapia Intensiva: un anno di esperienza.

9) SIRS in ICU: a different approach in sepsis assessment: clinical predictivity, severity scores, costs

1) A case of Hellp syndrome with a dramatic course successfully treated with early plasmapheresis.

R. Oggioni, GA Bocconi Minerva Anestesiologica 1993

A case of Hellp syndrome occurred in post-partum of a pregnancy elapsed without signs of preeclampsia is described. The evolution was particularly dramatic. Early plasmapheresis was the key treatment for a complete reco\-very in order to avoid Multiple Organ Dysfunction System (MODS). Emphasis on a multidis\-ciplinary approach between Gynecologist, Anesthesist, Nephrologist and Hematologist is pointed out.

Key words: Hellp syndrome - Plasmapheresis,

2.) Carbamazepine poisoning treated with plas-mapheresis. Case report.

A case of severe Carbamazepine poisoning in-itially misdiagnosed is reported. Treatment consisted in plasmapheresis (3.5 liters ex\-changed)repeated for 3 consecutive days, in conjunction with activated charcoal and advanced life support. It was obtained a rapid decay in Carbamazepine plasmatic level (with rebound phenomenon only after first treatment day) and a contemporary improvement in clinical conditions. The patient was discharged without complications after 6 days stay in ICU. Taking pharmacokinetic characteristics into account, it is suggested that plasmapheresis may be useful in this kind of poisoning.

Key words: Carbamazepine - Poisoning - Plas-mapheresis.

3) ACUBASE a new clinical tool for quality of care evaluation in Intensive Care

Bocconi G.A., Oggioni R., Tulli G., Palazzo M.* N.O. San Giovanni di Dio, Firenze.

*Charing Cross Hospital, London UK.

There is considerable interest in quality of care, severity of illness and their relationship to costs and resources in Intensive Care (ICU). We have been using a specific Intensive Care database program (AcuBase, CIS Seattle USA) for the last 8 months in which we have been able to document the following scores: SAPS, APACHE II, APACHE III, RTS, ISS, TRISS, 0SF, OFS, TISS with their associated calculations for risk of death and ICU costs.

AcuBase is a user friendly program that has allowed us to quantify our quality of care with standard international criteria. However it has also revealed a number of problems related to data acquisition in a general Italian hospital. These include the difficulty of determining the fate of patients once they have left the intensive care unit, so hospital mortality is difficult to quantify. Other problems are those of deciding on precise diagnostic categories, overscoring on Chronic Health Evaluation and errors in GCS in sedated patients who are cerebrally intact. These problems usually lead to an exaggerated risk of death estimation, a low Standard Mortality Ratio and false confidence in the quality of care offered by ICU. We would therefore recommend that data for scoring systems is collected by a senior doctor who is familiar with scoring systems and their literature. Our experience would also suggest that severity scoring should be done by a limited number of people to avoid discrepancies in interpretation. Overall, although we accept that there are some objections to scoring systems for prognostic evaluation of ICU patients, our view is that a spectrum of scoring systems such as offeredp by AcuBase together with one's own clinical assessment of probable patient outcome is a very educational and revealing tool for assessment of ICU efficacy, especially when related to costs.

234 MINERVA ANESTESIOLOGICA Ottobre 1993

4) PERCUTANEOUS DILATIONAL TRACHEOSTOMY (PDT) AND DIFFICULT AIRWAYS: IS FIBROTRACHEOSCOPY (FTS) ALWAYS ABLE TO AVOID COMPLICATIONS?

R.Oggioni, D.Gambi, G.Tulli

Intensive Care Unit Nuovo Ospedale San Giovanni di Dio - Firenze Italia

PDT is generally recognized as a safer technique than standard surgical tracheostomy. Nevertliless PDT is not free from early and late complications so it seems reasonable and useful to subject it to continuous criticals verifications. We have retrospectively anlized our record of 95 Ciaglia's and Griggs tracheostomies trying to assess as FTS could prevent some severes complications.

All PDT were performed by the same intensivists skilled in the procedure and in the fibrotracheoscopy.

FTS let us to perform dilational tracheostomy avoiding pararnedian tracheal puncture or vasculars lesions but in the same time let us to see how in many critically ill patients the tracheal mucosa was quite always ecchymotic or bleeding because prolonged endotracheal intubation (despite our PDT timing was 7 days). In our experience FTS was been absolutely necessary but not sufficient to prevent some traumatic lesions of the posterior tracheal wall.

We report two cases, one occurred in a 73 year old diabetic woman with a short neck admitted in ICU for severe sepsis tracheostomized after 9 days who developed after the end of procedure a sudden large subeutanequs emphysema of the neck, face, thorax caused by a posterior tracheal lesion which required a surgical repair through posterolateral thoracotomy without any subsequents consequence from this complication, the other one occurredin a COPD, vasculopathic patient and resulted in a esophageal fistula which has heavily conditioned his ICU length of stay.

The narrowing of tracheal stoma, peculiarity of PDT, and the initial lack of experience of some colleagnes at the first cannula changing, was the cause of the death in two cardiopathic patients (those complications occurred in the first PDT performed and before we established an accurate protocol of tracheal cannula changing which provides dilators and FTS use).

We believe that it is very important to point out those and above mentioned severes complications related in our opinion to "Difficult airways". Very likely, infact, there are previous or subsequents to PDT anatomopathologicals factors which on one hand make FTS essential to the procedure but on the other don't warrant the absence of complications.

Bibliography:

1. 1) P.Ciaglia et al. Percutaneouddilational tracheostomy. Results and follow up.CHEST 1992 ;lOl:464-67.
2. 2) W.B.Winkler et al. Bedside percutaneous dilational tracheostomy with endoscopic gnidance: experience with 71 ICU patients. INTENSIVE CARE MEDICINE 1994;20:476-79
3. 3) W.M.Griggs et al. A prospective comparison of percutaneous tracheostomy technique with standard surgical tracheostomy. INTENSIVE CARE MEDICINE 1991 ;17:261-3
4. 4) R. Oggioni et al. La tracheostomia percutanea dilatativa. Considerazioni critiche dopo un anno di sperimentazione. MINERVA ANESTESIOLOGICA 1995 (In press)

5) A COMPARISON BETWEEN A DYNAMIC( OFS) AND A STATIC (SAPS II) SEVERITY DISEASE INDEX

G.A.Bocconi, R.Oggioni ,G.Tulli

Intensive Care Unit Nuovo Ospedale San Giovanni Di Dio - Firenze Italia

As from the end of seventies, disease scoring system have been developed not only with the purpose to estimate the severity of illness in ICU patients and foresee the probabilty of death's risk, but even to verify the quality of care, auditing, resources monitoring and cost-benefit ratio. Those severity disease indexes (S.D.I.) ideally could be useful to decide on "Futility" and "Utility" of intensive care resources in some acutelly ill patients to limit an inappropriate ICU use. However S.D.I. don't reckon neither patient' sindividuality nor therapy or physiological reserves modifications and there is a considerable risk of false positive patients(from 10 to 20%); they are predicted dead but they live. Among S.D.I, some like APACHE and SAPS, are "statics", that's based on a single set of data recorded in the first 24 hours from TCU admission. Other, like M.P.M. and O.F.S. are "dynamic" scoring systems, because data are collected day by day after ICU admission and reckon developing of MSOF. But pathologicals processes in ICU patients are dynamic so the aim of our study was been to confront two SDI,one static SAPS II, the other one dynamic OFS, to verify their predictives capacities on an acknowledged endpoint, the ICU mortality. Among 152 patients admitted in our ICU from 1-1-94 to 31-3-95 we selected retrospectively 31 patients in whom we correlated the two SDI. SAPS II and APACHE II, necessary for OFS formula, were automatically calculated from our computerized ICU prograrn ACUBASE 3.4 version (C.I.S. Seattle-USA). OFS was recorded after 24,48 hours and until seventh day after admission. The results direct ourselves to the conclusion that OFS, a scoring system which follows time's progression of disease's severity in critically ill patients, shows not only a specifity of 91% as to 18% of SAPS TI but even, and this is more important, a correct value of predictivity of ICU mortality of 96% as to 71% of SAPS II. ROC Curves shows that the OFS area is equal to 93.15% as to 73% of SAPS II area. Therefore there is a better discrimination for OFS versus SAPS II.To confirm or not the superiority of OFS will be necessary a larger sample of patients, but certainly a scoring system which"takes" the patients during his ICU stay is more reliable than a scoring iystem base on data collected ni an only moment.

BIBLIOGRAPHY

1. 1)Chang R.W,Jacobs s.predicting out come among intensive care unit patients using computerised trendanalysis of daily Apache II scores corrected for organ system failure. INT. CARE MED. 14:558,1988
2. 2) LeGall JR, Lemeshow 5. A new simplified acute physiology score(SAPS II) based on an European North American Multicenter study. JAMA 270:22 Dec. 1993(n.24)

6) COMPARISON OF COMPLICATIONS BETWEEN PERCUTANEOUS AND TRANSLARYNGEAL TRACHEOSTOMY: A 3 YEARS EXPERIENCE.

G.Tulli, D.Gambi, R.Oggioni, M.Librenti

Intensive Care Unit, Nuovo Ospedale San Giovanni Di Dio, Firenze, Italia

INTRODUCTION In the last years percutaneous techniques of tracheostomy as described by Ciaglia, Griggs, and others have widely replaced the classic surgical operation in Intensive Therapy Unit (ITU) setting. A great deal of papers stress that new percutaneous techniques (PDT) offer many advantages especially a marked reduction of infectious risk of the wound, a more rapid favourable rate of spontaneus healing of the stoma and an overall containment of costs. More recently a new, original approach to the tracheostomy (translaringeal tracheostomy, TLT) has been proposed to overcome some problems the other PDT have still to face with. We report the results of a retrospective review of 153 consecutive tracheostomies (PDT and TLT) performed in these last 3 years in our ITU, in order to compare the complications rate of these techniques.

METHODS 153 consecutive patients, 96 males and 57 females, aged 66 ± 14 years, admitted for medical (73), surgical (63) or traumatic (17) pathology. The APACHE III on entry was 80 ± 26. The long of stay was 28 ± 25 days. Tracheostomy was performed early (6 ± 4 days) in each group, by the same 5 doctors' team. Fibrotracheoscopy (FTS) was used in 90% of PDTs.

RESULTS Major advantages of PDT are minimal skin incision (and thus an all but invisible scar), a substantial reduction of infectious complications of the wound, the bedside feasibility of the procedure, a simplified care of the stoma and a cost containment related to a lesser personnel request. In our series the problems we encountered in performing PDT were generally of minor concern: some transient difficulties in the dilation step, owing to collapse of the tracheal wall and a certain ease in the formation of tracheal flaps. All these problems are almost always preventable by a recognition, so a fibrotracheoscopic (provided by an another operator) supervision is needed all along the procedure. Four major postoperative complications that FTS failed to prevent were: 1 severe bleeding at the site of incision, 1 injury of posterior wall, 1 tracheoesophageal fistula requiring surgical repair and 2 stenosis by granuloma in malacic tracheas (1 of which unsuccessfully managed by laser therapy), probably caused by the dilation trauma itself. 3 patients suffered major complications during the first change of the cannula: 1 rupture of a tracheal ring, 1 wrong route and 1 cardiac arrest due to severe hypoventilation. In this patient several attempts of cannula replacement failed and no other successful kind of ventilation could be implemented. 4 additional cases of difficult replacement required a new short dilational procedure. TLT adds some further advantages in respect of PDT, e.g. less tissue trauma, since any injuries to the vessels of the neck and to the posterior wall of trachea are prevented (Fig. 1). Nevertheless potential drawbacks and hazards still exists and need a little higher degree of skillness and dexterity to be overcome or prevented. Moreover TLT requires that the endotracheal tube is replaced by a rigid tracheoscope (Fig. 2), a new intubation is needed and an optimal neck extension and mouth opening are of paramount importance. Rigid tracheoscope may be replaced by a fiberoptic tracheoscope (Fig. 3) but in this way one misses the opportunity of pushing up and protruding the anterior tracheal wall before the exploring needle is inserted. During the insertion and rotation of the cannula (Fig. 4) a high peak pressure is needed to ventilate the patient via the small internal diameter (4,5 mm) tube used. This step generally lasts for 2 or 3 minutes and a stop of the ventilation in the meantime could expose the patient to increasing hazards if a rescue system is not in use. In effect we had 1 case of severe desaturation in an ARDS patient during this manoeuvre. The correct positioning of the tracheal cannula can be difficult; it can require several attempts, though with increasing experience, this step becomes smoother. In only 4 cases we failed and the rotating and positioning of the tracheal cannula required a conversion into a dilational procedure. We had no postoperative complications. During first cannula changing we encountered difficulties in recannulating 2 patients and this required a new brief dilational procedure.

CONCLUSION Both PDT and TLT offer significant advantages over surgical tracheostomy and can gain further widespread acceptance in the ITU setting, provided that some mandatory rules are followed. In particular a defined training program, a clearly established nursing protocol, and FTS supervision all along the procedure are mandatory conditions. TLT is a slightly more complex procedure than PDT and a previous good experience with the dilational technique is advisable in order to shorten the learning time and to overcome unexpected difficulties without exposing the patient to any risk. On the other hand TLT may allow a further reduction in the rate of late, non FTS preventable, complications related to even mild injuries in malacic tracheas. An improvement in design and consistence of the cannula might help in making easier the insertion step and the aspiration manoeuvre. The first replacement of the cannula (when the stoma is fresh and can easily collapse) should be performed with FTS and dilational equipment at hand, using an "over the wire" technique.

7) P29 SIRS in ICU: a different approach in sepsis assessment: clinical predictivity, severity scores, costs

R Oggioni, GA Bocconi, V Mangani, Mascii, E Messeri and G Tulli

Intensive Care Unit, Nuovo Ospedale San Giovanni Di Dio, Via Torregalli 8, Firenze 50100, Italy Crit Care 1997, (suppl 1):P29\

The systemic inflammatory response syndrome (SIRS) based on the changes of four physiological features like temperature, white cells, heart rate and ventilation, can be observed after a wide variety of insults.

Because sepsis is the systemic response to infection and it is the most common cause of death and of multiple organ failure (MOF) in ICU, we have tried to use SIRS as a predictive tool against the risk of sepsis, severe sepsis and septic shock.

Many authors choose two of the four criteria of SIRS to verify this hypothesis, but results were misleading; a significant predictive power of SIRS against sepsis was not found.Although SIRS is not disease specific, we performed retrospective study on 384 unselected patients admitted consecutively to our ICU from 1 January 1993 to 31 May 1995, connecting epidemiological data to our research.

We found that three-four criteria of SIRS (group SIRS 1) are significantly better than the two criteria pattern (group SIRS 2) as predictive power.

This result was confirmed by the significant difference of APACHE III and SAPS II scores (87 and 55 respectively versus 78 and 48), by the longer length of stay of survivors (24 days versus 12) and by bigger costs (78 million versus 38 million italian lire) in the group SIRS 1.

We concluded that by adding one or two SIRS criteria to those normally used, the predictive liability of SIRS against risk of sepsis is significantly enhanced.

8) VALUTAZIONE CLINICA DEL DOSAGGIO DEGLI ANTIBIOTICI AMINOGLICOSIDICI E GLICOPEPTIDICI IN TERAPIA INTENSIVA: UN ANNO DI ESPERIENZA.

A. Veneziani, R.Oggioni, R.Iamello, F.Tonelli*, G.Tulli

U.O. Anestesia e Rianimazione, *U.O. Laboratorio e Analisi

INTRODUZIONE

Il dosaggio ematico degli antibiotici aminoglicosidici e glicopeptidici impiegati in terapia intensiva è particolarmente importante nei pazienti critici. Per mantenere concentrazioni plasmatiche terapeuticamente efficaci di questi antibiotici è difficile stabilirne a priori la dose adeguata che spesso può risultare aumentata o ridotta rispetto ai dosaggi standard basati sul peso corporeo del paziente. Quindi un monitoraggio regolare delle concentrazioni plasmatiche potrebbe esser utile in virtù della nota tossicità di questi farmaci che rende assai esiguo il margine tra rischi e benefici.

Lo scopo del nostro lavoro è stato quello di verificare clinicamente quanto il dosaggio ematico di tali antibiotici, si sia rivelato adeguato a mantenere entro un corretto range terapeutico la posologia e se quindi tale metodica possa essere introdotta routinariamente nel nostro reparto.

MATERIALI E METODI

E’ stato condotto uno studio retrospettivo su n° 47 pazienti ricoverati presso la terapia intensiva dal 1 aprile 1994 al 31 dicembre 1995 ai quali sono stati somministrati uno o più degli antibiotici oggetto del nostro studio. Criteri d’inclusione erano un numero di dosaggi ematici per singolo antibiotico non inferiore ai 3 campionamenti. Di ogni paziente è stata riportata l’età, il sesso, il peso, la diagnosi d’ingresso, il SAPS II, l’APACHE III, la creatininemia del giorno d’ingresso, la lunghezza della degenza, l’outcome.

Il dosaggio degli antibiotici è stato limitato ad Amikacina, Tobramicina per gli aminoglicosidici e alla Vancomicina per i glicopeptidici.

La posologia e l’intervallo delle somministrazioni degli antibiotici erano stabiliti giornalmente su base clinica, tenendo conto per la prima somministrazione dell’età, del peso del pz. della sua patologia, di altre terapie antibiotiche e non associate, dell’output urinario, di eventuale terapia sostitutiva renale (CVVH o HD). Per le somministrazioni successive invece sono state considerate anche le concentrazioni di valle e di picco di ciascun antibiotico. I campioni ematici sono stati analizzati con la metodica dell’immunofluorescenza (Tdx Abbott Laboratories) presso il Department of Haematology and Clinical Chemestry of our hospital.

Il dosaggio serico degli antibiotici è stato effettuato una volta al giorno 3 volte alla settimana con modalità rigorosa: campionamenti 15 minuti prima della somministrazione del farmaco (concentrazione di valle) e 5-15 minuti o 60-120 minuti dal termine dell’infusione (concentrazione di picco) rispettivamente per gli antibiotici aminoglicosidici e glicopeptidici

Sono stati inoltre identificati nell’ambito della popolazione studiata due gruppi, pazienti non dializzati (Gruppo N.D.) e pazienti con insufficienza renale trattati con un supporto sostitutivo (Gruppo D), e ciò per verificare se vi fosse una diversa rilevanza del dosaggio antibiotico sulla base di evidenti alterazioni dei tassi di creatininemia.

Per ogni paziente e per ogni singolo antibiotico dosato è stato annotato il n° di giorni di terapia, il n° dei giorni in cui è stato dosato, espresso anche come valore percentuale (media ±SD), il dosaggio medio giornaliero e i valori dei livelli di valle e di picco degli antibiotici (media ponderata ±SD). Sono stati registrati il valore minimo e max dei valori di picco e di valle riscontrati in ogni paziente espressi (media ± SD) ed il n° di volte in cui le concentrazioni di valle e di picco risultavano minori o maggiori del 20% rispetto ai limiti del range teorico stabilito dei singoli antibiotici (Amikacina concentrazione di valle 5-20 mcg/dl, picco 20 -35 mcg/dl; Tobramicina valle 1-2 mcg/dl picco 3-14 mcg /dl ; Vancocina Valle 10-20 mcg/dl, Picco 30-40 mcg/dl) Tali valori sono stati scelti come rappresentativi di un sicuro errore posologico (numero di volte, percentuale, mediana, range).

Sono state infine annotate il numero di volte in cui sulla base del dosaggio ematico eseguito vi è stato un aggiustamento della posologia (valore assoluto e percentuale sul numero di somministrazioni).

I dati riportati sono riferiti all’intero gruppo studiato (Dati Globali) ed ai due sottogruppi (N.D e D). L’analisi statistica ha riguardato, quando il numero dei casi lo ha consentito, il confronto tra questi due sottogruppi. E’ stato utilizzato il test t di Student per dati non appaiati per ciò che concerne i dati relativi alla popolazione e il dosaggio, la durata della terapia, il numero dei giorni di dosaggio ed il loro rapporto, le dosi, le concentrazioni di valle e di picco e le concentrazioni massime e minime di valle e di picco rilevate. Per il confronto dei dati relativi allo sfondamento del ± 20% del range terapeutico delle concentrazioni ematiche di valle e di picco, i singoli valori dei due gruppi sono stati analizzati con il test u di Mann Whitney. Infine il confronto sempre fra i due gruppi N.D e D fra le variazioni di posologia effettuate sulla base del campionamento ematico è stato utilizzato il test del chi -quadro. In tutti i casi è stata considerata statisticamente significativo una p<0,05.

RISULTATI

La popolazione da noi analizzata è risultato composta da n° 32 uomini e da n° 15 donne . Ad essa si riferisce la Tab 1 . La causa di ammissione in ICU è risultata da 7 politraumatismi, 12 sindromi emorragiche postoperatorie, 15 insufficienze respiratorie, 2 ARDS, 5 sepsi, 1 embolia polmonare, 2 sindromi postanossiche da arresto cardiaco, 2 scompensi cardiaci congestizi, 1 stato di male epilettico.

Nelle tabelle successive i dati sono riferiti a tutti i pazienti (Dati Globali) ed ai 2 gruppi ND e D).

Nella Tab. 2 e 3 sono riportati i dati inerenti l’Amikacina: la Tabella 2 è riferita ai giorni di somministrazione e dosaggio dell’antibiotico e al rapporto tra questi due (% ratio), alle dosi ed ai valori di valle e di picco di detti dosaggi (media ponderata e SD). Nella Tabella 3 sono riportati i valori massimi di valle e di picco, il numero di volte che il range terapeutico è stato sfondato del ±20% dei suoi limiti inferiore e superiore (espresso anche come mediana con il range dei valori riscontrati).

Analogamente la Tab. 4 e 5, 6 e 7, rispettivamente, riferiscono gli analoghi dati concernenti il dosaggio di, Tobramicina e Vancomicina.

Nella Tabella 8 infine è riportato per ogni singolo antibiotico il numero di volte che in base ai dati di laboratorio è stata modificata la posologia, dato espresso sia in percentuale dei pazienti che come percentuale rispetto al numero totale di dosi somministrate.

Tabella 1

Amikacina tab 2

* P= 0.93 N.S. ** P = 0,11 N.S.

Tab 3

Tab 4 Tobramicina

* P = 0,92 **P = 0,49

Tabella 5

Tabella 6 Vancocina

* P = 0,29 ** P= 0,27

Tabella 7

Tab. 8 Variazioni della posologia sulla base del dosaaggio

DISCUSSIONE

Mantenere il tasso ematico degli antibiotici all’interno di un coretto range terapeutico è indubbiamente non semplice. Sono troppe le variabili in gioco per ogni paziente: età, peso, volume di distribuzione del farmaco, patologia, funzionalità renale, tipo di antibiotico impiegato, concomitanza di altre terapie antibiotiche e non, per poter stabilire a priori la posologia giornaliera in maniera standard Indubbiamente la misura della concentrazione di un farmaco permette al medico di adattare al singolo paziente (to tailor) il dosaggio e di ottenere il massimo effetto terapeutico con il minor rischio di tossicità.

Il metro di giudizio più appropriato è quindi sicuramente quello clinico, che è in grado di mediare le varie informazioni traducendole in una prescrizione. Il dosaggio degli antibiotici , in tal senso è un’utilissima informazione in più a disposizione del medico per attuare ed in seguito verificare la posologia impiegata. Dai nostri dati emerge che pur attuando tale metodica con precisione , gli errori di sovra e sottodosaggio farmacologico sono tutt’altro che infrequenti per cui a nostro giudizio la tecnica si prospetta come un elemento importante per correggere errori posologici non altrimenti riconoscibili col solo giudizio clinico.

SIRS in ICU: a different approach in sepsis assessment: clinical predictivity, severity scores, costs

R.Oggioni, GA. Bocconi, V. Mangani, F. Mascii, E. Messeri and G. Tulli

Intensive Care Unit Ospeale Nuovo San Giovanni di Dio Firenze

INTRODUCTION

Nosocomial infections are the most frequent complications in ICU patients.

Major risk factors for ICU patients are: extreme age, severity of pathology on entry, previous surgicals operations, mechanical ventilation, gastric alkalyzation, invasive procedures, organ dysfunctions, decrease of immunity, state of nutrition, broad spectrum antibiotics, multiresistant microganism, length of stay (1-2-3-4-5-6-7). When the infection is associated to a systemic inflammatory response (SIRS) we have SEPSIS which represents one of the major cause of morbidity and mortality in critical ICU patients (8-9-10-11-12).

Sepsis implies both a rise of direct costs related to the increased length of stay and a rise of indirect costs for patients, families and society. (13-14).

Since 01-01-93 in our ICU we have systematically stored data on infections (site of isolation, germs) on a specific ICU software called ACUBASE 3.4 â (15).

Our data were subsequently epidemiologically analized and became a subject of discussion, an audit of our procedures and behaviour.

With it we carried out our research with the aim to verify whether Sirs, defined by more than two criteria had a major predictive power towards sepsis than Sirs with two criteria.

Afterwards we considered:

• 1) The incidence of sepsis and how many ICU patients developed sepsis within the first seven days of stay and how many after.
• 2) The outcome, the influence of age and length of stay of patients with Sirs and sepsis.
• 3) The significance of the more common severity scores (SAPS II and APACHE III) in Sirs and sepsis patients. (16-17).
• 4) The TISS of Sirs and sepsis patients for costs estimations (18).

PATIENTS AND METHODS

We carried out a retrospective study on 384 patients admitted in the ICU of Nuovo Ospedale San Giovanni di Dio in Florence, from the 1st of January 1993 to 31st of May 1995. Our ICU is a general seven bedded ICU with a nurse/patient ratio of 1:2.

The patients have been registered on the basis of age, sex and pathology on entry

TABLE I : POPULATION CHARACTERISTICS AND PATHOLOGY ON ENTRY

We have selected all the patients conforming to definition of Sirs and sepsis suggested by the Consensus Conference of American College of Chest Physicians and Society of Critical Care Medicine (19):

SIRS (Systemic Inflammatory Response Syndrome). The response is manifested by two or more of the following conditions: temperature > 38°or < 36°; white blood cell count >12.000/cu mm or < 4000/cu mm; heart rate > 90 beats per minute; respiratory rate > 20 breaths per minute or PaCO2 < 32 mm Hg .

SEPSIS: The systemic response to the presence of a confirmed infection

SEVERE SEPSIS: Sepsis associated with organ dysfunction (MODS), hypotension (< 90 mm hg) or hypoperfusion (oliguria, alterations of GCS, lactic acidosis. MODS (multiple organ dysfunction syndrome): presence of altered organ function in an acutelly ill patient such that homestasis cannot be mantained without intervention.

SEPTIC SHOCK: Sepsis associated to hypotension persisting despite adequate fluid resuscitation, with presence of perfusion abnormalities. The patients need inotropic or vasopressor drugs to mantain arterial pressure.

Patients with Sirs were divided out in two groups:

A) Group SIRS 1: Patients with three or four criteria of Sirs (see above)

on entry: alterations of temperature, of white blood cells, of heart rate, of respiratory rate or presence of mechanical ventilation.

B) Group SIRS 2: Patients with two criteria of Sirs on entry.

We have considered within these two groups the presence (from the first to the fourteenth day of stay) and the precocity (within the first seven days of stay) of sepsis, severe sepsis and septic shock.

For the severe sepsis we considered the presence of at least two organ dysfunctions, p.e. a PaO2/FiO2 < 250 mm Hg and an arterial pressure < 90 mm Hg (20-21).

For the septic shock we chose those patients with hypotension (BP < 90 mm hg), oliguria (urinary output < 0.5 ml/kg /h), who, despite adequate fluid resuscitation, needed vasoactive drugs like Dobutamine > 10 mcg/kg or Norepinefrine > 0.1 mcg/kg.

We also pointed out the presence of the infections without Sirs (TABLES II-III).

For statistical comparison between the two groups we used Chi square test with p < 0.03 (TABLE IV). In succession we reported within the two groups of Sirs ,the incidence and the outcome of sepsis, severe sepsis and septic shock (TABLE V), the length of stay (TABLE VI), the Tiss and mean costs of patients survived (TABLE VII) , the severity scores to the admission in ICU and in the day of infection (TABLE VIII) , the risk of death related to the age (TABLE IX).

We have tried to exclude the patients with bacterial colonization: in order to achieve this we considered only the samples with a well documented cfu/ml (colony forming unit/ml), indicating a level of infection, p.e. a cfu of 10 ^ 6 for tracheal aspirates, a cfu of 10 ^ 4 for BAL (broncoalveolar lavage), a cfu of 10 ^ 3 for PSB (protected specimen brush), a cfu > 15 for intravascular catheters (22-23).

The microbiological survey was carried out by submitting the patients, on entry in ICU to the drawing of the pharingeal sample, sputum or tracheal aspirate, blood cultures, urinoculture, previous intravascular catheters, surgical deep wounds etc. Afterwards the patients underwent surveying routine cultures twice a week or more, if dictates by clinical course.

In case of dubious or negative isolation from tracheal aspirate we draw protected or guided specimens of lower respiratory tract.

RESULTS (see Tables II-IX)

The patients with Sirs to the admission in ICU were 259 (67%).

TABLE II : GROUP SIRS 1 ( 3-4 criteria )

TABLE III: GROUP SIRS 2 ( 2 criteria )

TABLE IV: COMPARISON BETWEEN THE TWO GROUPS OF SIRS

Sepsis are reported in the first and third column, Sirs in the second and fourth.

TABLE V: SEPSIS-MORTALITY (Whole population)

** Total sepsis are 117 because 9 patients developed sepsis after the 14th day of ICU stay.

(117 - 9 = 108).

TABLE VI : LENGTH OF ICU STAY

(LOS/days) Mean + SD

TABLE VII: TISS / LOS SURVIVED COSTS (Italian lire)

TABLE VIII: SAPS II AND APACHE III SIRS 1 AND SIRS 2 GROUPS

TABLE IX :SIRS 1 AND SIRS 2 AGE/MORTALITY

(years in deciles)

DISCUSSION

Our case mix shows more medical than surgical or traumatic patients.

We find few trauma cases in our unit because they are sent to university hospital even though they are not trauma center.

Comparison between Sirs 1 and 2 groups displays fewer patients but more predictive power in four criteria group versus the two criteria one.

In the former 54% developes sepsis, 17% severe sepsis and 12% septic shock: in the latter one 38%, 4%, 2% respectively.

Sepsis in the first week was found in 68% of Sirs 1 patients versus 47% in Sirs 2.

Statistically these are significant data: chi square test is equal to 10.263 with p< 0.03.

Mortality was 52% in Sirs 1 group, 47% in Sirs 2 group (Tables II, III, IV).

At the admission we found a rate of Sirs lower (67% versus 80%) than reported in two studies by Pittet and coll. (24-25).The Italian SEPSIS Study Group (26) reported a rate of Sirs of 52%.

In the above mentioned studies of Pittet the case mix is very different because author’s ICU admits exclusively surgical patients. Moreover sepsis rate reported is also higher than in our study (49% versus 30%): on the other hand sepsis is defined by Pittet only by administering empirical antibiotics for three days as opposed germ identification as was performed in our study. A higher number of sepsis are reported in the same study after discharge from ICU: it means that many patients are discharged not necessarily recovered.

There were 117 sepsis (9 of them after the second week of stay and not included here): 41 (35%) were surgical, 68 (58%) medical and 8 (7%) in trauma patients (Table V).

Our sepsis mortality rate, in hospitalized patients, was 33%, the lenght of stay, referred to ICU, 14 days.

Tables VI, VII, VIII respectively process ICU length of stay, staff burden (basing on Tiss

score), severity disease indexes (SAPS II and APACHE III) in the Sirs group; we remark a clear difference between SIRS 1 and SIRS 2 groups (higher and lower rates respectively).

Referring to Table VII, we point out a greater total cost in SIRS 1 survivors than in the SIRS 2 group (more than 50%). Costs are calculated basing on a mean cost of Tiss point equaling to 57.000 Italian lire (27-28).

The higher SAPS II and APACHE III in SIRS 1 group on entry affect the mortality rate and the longer length of stay of those patients.

In SIRS 1, mortality rate is three times greater than in SIRS 2 group: this difference increases with age (Table IX).

SITE OF INFECTION AND ISOLATED GERMS (See Infections Graphs)

Processing microbiological data we find 41% of infections in the lower respiratory tract; 37% of them are pneumonias. 50% of patients admitted in our ICU shows previous lung infections.

There is a clear prevalence of gram positive germs against gram negative (49% vs 39%). Gram positive germs were : Staphylococcus aureus (20%), Enterococcus (15%), Staphylococcus coagulasis-negative (10%), Streptococcus pneumoniae (5%). Most staphylococcus aureus strains (90%) are meticilline resistant.

Gram negative germs were Enterobacteriacee (Klebsiella, Escherichia) (12%), Pseudomonas aeruginosa (7%), Serratia marcescens (7%), Xanthomonas maltophilia (6%).

Fungi (mainly Candida spp.) were 12%.

We have had a relatively high rate (11%) of deep surgical wounds (surgical abdominal drain); this was related to a 20% of further surgical procedures in patients with a severe abdominal sepsis.

CONCLUSIONS

Sepsis is one of the most common cause of death in ICU and the main cause of multiple organ failure (MOF).

Because its monitoring and therapy are very laborious and time consuming for the staff in ICU, unconfortable for patients and expensive for families and the society, it' s clear that a good predictive tool for it would be welcomed.

Previous trials based on SIRS (Systemic Inflammatory Response Syndrome) gave poor results as far as the prediction of sepsis; is concerned the main reason was the fuzziness of criteria defining SIRS. variations of white cells, temperature, heart and respiratory rate are at all aspecific and common among several diseases.

Furthermore the use of only two casual criteria produces a much weaker predictive power of Sirs and sepsis development.

Processing our data, which cannot be compared to other studies because of the different case mix, highlights a statistically significative difference between four criteria SIRS and two criteria analogue.

SIRS 1 group shows a predictive power of sepsis development clearly better than SIRS 2 group.

According to our research it would seem to be more useful to modify the definition of SIRS proposed by the Consensus Conference ACCP/SCCM, with a four criteria definition.

Microbiological data show that 70% of respiratory infections were acquired in the first week of stay in ICU; these infections affect the large majority of SIRS 1 group patients.

The other infections (urinary, surgical wounds) were acquired also during the first week of stay but affect the SIRS 2 group.

Line and blood stream infections were acquired after the first week and affect most of all patients of SIRS 2 group.

Gram positives represent the majority of germs. Perhaps this prevalence is in relation to the complete turnover of the nursing staff with the introduction of many new qualified younger nurses; so much time was spent in their training and for the accomplishment of the care protocols concerning prevention and control of infections .

The rate and severity of nosocomial and ICU infections can be diminished only by effective prevention and control measures like careful and scrupulous hygiene (handwashing above all), correct and aseptic invasive procedures, frequent change of intravascular devices, of ventilator circuits and humidifiers, correct patient's positioning, appropriate bronchoaspiration, closed ventilation suction system, correct antibiotic therapy, as well as a well working microbiological laboratory (29-30).

BIBLIOGRAPHY

1) Jarvis WR, Martone WJ. Predominant pathogens in hospital infections. Hospital infections program, CDC, Atlanta, Georgia, USA. Journal Antimicrob Chemiother 1992; 29 (Supl.A): 19-24 2) Inglis TJ, Sproat LJ, Hawkey PM, Knappet P . Infection control in intensive care units: UK national survey 1990. Br J Anaesth 1992 68: 216-220 3) Spencer RC. Epidemiology of infections in ICUs. Intensive Care Med 1994 20:S2-S6 4) Widmer AF. Infection control and prevention strategies in the ICU. Intens Care Med 1994 20: S7-S11 5) Donowitz LG, Wenzel RP, Hoyt JW. High risk of hospital acquired infections in the ICU patients. Crit Care Med (1982) 10:355-357 6) Chevret S, Hemmer M, Carlet J, Langer M and the European Cooperative Group on Nosocomial Pneumonia(1993) Incidence and risk factors of pneumonia acquired in intensive care unit. Results from a multicenter prospective study on 996 patients. Intens Care Med 19:256-264 7) Bjerke HS, Leyerle B, Shabot MM. Impact of ICU nosocomial infections on outcome from surgical care. Am Surg 1991;57:789-802 8) Bone RC Toward an epidemiology and natural history of SIRS (systemic inflammatory response syndrome). JAMA (1992)268:3452-3455 9) Tulli G. La sepsi. Mcgraw-Hill Italia srl 1995 p.9-15 10) Craven DE, Kunches LM, Lichtenberg DA, Kollisch NR, Barry MA, Heeren TC, McCabe WR (1988). Nosocomial infection and fatality in 4edical and surgical intensive care units. Arch Intern Med 148:1161-1168 11) Wenzel RP. The mortality of hospital- acquired bloodstream infections: need for a new vital statitics? Int J Epidemiol 1988;17:225-227 12) Fagon JY, Novara A, Stephan F et al. Mortality attributable to nosocomial infections in the ICU. Infect Control Hosp Epidemiol1994;15:428-434 13) Spengler RF, Greenough WB. Hospital costs and mortality attribuited to nosocomial bacteremias. JAMA 1978:240:2455-2458 14) Garo B, Boles JM. Le cout de l'infection nosocomial en reanimation: une evaluation medicale et economique. Rean Urg 1993;2:648 15) Tulli GA, Bocconi GA, Oggioni R et al: Acubase: fundamentals and perspectives. APICE 10° Postgraduate Course in Critical Care Medicine. Trieste 13-19/11/95 p.641-651 16) Knaus WA,Wagner DP, Draper EA, Zimmerman JE et al. The APACHE III Prognostic System: risk prediction of hospital mortality for critically ill hospitalized adults. Chest (1991) 100:1619-1638 17) LeGall JR, Lemeshow S. A new simplified acute physiology score (SAPS II) based on European North American Multicenter Study. JAMA December 1993 Vol. 270 N°24; 2957 18) Keene AR, Cullen DJ. Tiss:Therapeutic Intervention Scoring System-Update 1983. Crit Care Med (1983) Vol 111 N°1: 1-3 19) American College of Chest Physician/Society of Critical Care Medicine Consensus Conference: Definition for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med (1992) 20:864-874 20) Fagon JY, Chastre J, Novara A, Medioni P and Gilbert C. Characterization of intensive care unit using a model based on the presence or absence of organ dysfunctions and/or infection: the ODIN model. Intens Care Med(1993) 19:137-144 21) Hebert PC, Drummond AJ, Singer JS, Bernard GR, Russel JA (1993). A simple multiple system organ failure scoring system predicts mortality of patients who have sepsis syndrome. Chest 104:230-235 22) Langer M, Pifferi S, Peta M. Diagnosis of bacterial infection in ICU: general principles. Intens Care Med (1994) 20:S12-S16 23) Hospital Acquired Pneumonia in adults: Diagnosis, assesment of severity, initial antimicrobial therapy and preventives strategies. A consensus statement. Am J Resp Crit Care Med 1995 Vol. 153:1711-25 24) Pittet D, Rangel-Frausto S, Li N, Tarara D, Costigan M et al. Systemic inflammatory response syndrome, sepsis, severe sepsis and septic shock: incidence, morbities and outcome in surgical ICU patients. Intens Care Med(1995) 21:302-309 25) Rangel-Frausto MS, Pittet D, Costigan M et al. The natural history of the Systemic Inflammatory Syndrome (SIRS). A prospective study. JAMA 1995: 273:117-123 26) Salvo I, De Cian W, Musicco M, Langer M, Piadena R, Wolfler A, Montani C, Magni E and The SEPSIS Study Group. The Italian SEPSIS study: preliminary results on the incidence and evolution of SIRS, sepsis, severe sepsis and septic shock. Intens Care Med (1995) 21:S244-S249 27) Kappstein I, Schulgen G, Beyer U et al. Prolongation of hospital stay and extra costs due to ventilator-associated pneumonia in an intensive care unit. Eur J Clin Microbiol Infect Dis 1992:11:504-508 28) Cowley JM, Hill S. Handwash practice in ICU: the effect of an educational program and its relationship to infections rate. Am J of Infection Control 1989,17:330-339 29) Moro ML, Jepsen OB. The EURO.NIS STUDY GROUP. Infection control practices in intensive care units of 14 European countries. Intens Care Med (1996) 22:872-79