ARTICLE IN PRESS

Respiratory Medicine (2006) 100, 529–535

KEYWORDBronchoalvlavage;BronchoscoDiagnosis;Complicati

0954-6111/$ - sdoi:10.1016/j.r

Abbreviationexpiratory volu

$Funded by�CorrespondiE-mail addr

1The other mtubing) who paMaria Teresa MDepartment ofFelipe Coll, ReDepartment ofHospital de Sab(Girona), CarleDepartment ofRibes (TarragonRespiratory Me

A new BAL fluid instillation and aspirationtechnique: A multicenter randomized study$

Antoni Rosella,�, Antoni Xaubetb, Carles Agustıb, Joan Castellac,Carmen Puzoc, Vıctor Curulld, Javier de Graciae,on behalf of the RASTA study group1

aServei de Pneumologia, Respiratory Medicine Department of Hospital Germans Trias i Pujol,Carretera de Canyet s/n. 08916 Badalona, Catalonia, SpainbRespiratory Medicine Department of Hospital Clınic, IDIBAPS, Barcelona, Catalonia, SpaincRespiratory Medicine Department of Hospital de Sant Pau, Barcelona, Catalonia, SpaindRespiratory Medicine Department of Hospital del Mar, Barcelona, Catalonia, SpaineRespiratory Medicine Department of Hospital Vall d’Hebron, Barcelona, Catalonia, Spain

Received 24 January 2005; accepted 31 May 2005

Seolar

py;

ons

ee front matter & 2005med.2005.06.002

s: BAL, bronchoalveolame in the first second;the Societat Catalana dng author. Tel.: +34 93ess: trosell@ns.hugtip.sembers of the RASTArticipated in this workelgosa, Respiratory MeHospital del Mar (Barcespiratory Medicine DeHospital de Sabadell—adell—Corporacio Parcs Sabadell, RespiratoryHospital de St. Pere de Ra), Jaume Oriol, Respdicine Department of H

Summary Bronchoalveolar lavage (BAL) is a widely used clinical and research tooleven though certain steps in the procedure have not been standarized. One step thatis subject to variation is the technique used to instill and recover BAL fluid. Our aimwas to assess whether attaching a plastic tube between the syringe and the workingchannel of the flexible bronchoscope allowed recovery of more fluid in comparisonwith recovery using only a syringe, and whether tubing had a favorable impact ondiagnostic yield and complications.

Elsevier Ltd. All rights reserved.

r lavage; CI, confidence interval; COPD, chronic obstructive pulmonary disease; FEV1, forcedSD, standard deviation; SpO2, arterial oxygen saturation by pulse oximetrye Pneumologia (SOCAP) and Red Respira (ISCiii-RTIC-03/11)497 89 20; fax: +34 93 497 88 43.cs.es (A. Rosell).study group (Rentat Amb o Sense Tubuladura Associada—Lavage with or without associatedare: Elena Riu, Respiratory Medicine Department of Hosptial Germans Trias i Pujol (Barcelona),dicine Department of Hospital Clınic (Barcelona), Maria Jose Santos, Respiratory Medicinelona), Edelia Catalan, Respiratory Medicine Department of Hospital Vall d0Hebron (Barcelona),partment of Hospital de la Creu Roja (Barcelona), Miguel Gallego, Respiratory MedicineCorporacio Parc Taulı, Sabadell (Barcelona), Elsa Naval, Respiratory Medicine Department ofTaulı, Sabadell (Barcelona), Rosa Jolis, Respiratory Medicine Department of Hospital de FigueresMedicine Department of Hospital de Figueres (Girona), Jesus Ribas, Respiratory Medicineibes (Tarragona), Josep Maria Vidal, Respiratory Medicine Department of Hospital de St. Pere deiratory Medicine Department of Hospital Municipal de Badalona (Barcelona), Felipe Andreo,ospital de Sant Boi (Barcelona).

ARTICLE IN PRESS

A. Rosell et al.530

A randomized multicenter study was performed in eleven public hospitals inCatalonia (Spain). Patients scheduled for BAL were randomly assigned to tubing(n ¼ 140) or no-tubing groups (n ¼ 155). Flexible bronchoscopy and a 150mL BAL inthree 50mL aliquots with or without tubing attached to the syringe.

In the group with tubing, 8% more fluid was recovered. The difference was bothstatistically significant and clinically relevant given that 17.4% more diagnoses, 6.9%fewer complications, and 8.8% fewer technical failures were recorded.

Based on these results, we recommend performing BAL using plastic tubingbetween the 50mL syringe and the working channel of the flexible bronchoscopewhen carrying out manual instillation and suction.& 2005 Elsevier Ltd. All rights reserved.

Introduction

Bronchoalveolar lavage (BAL) is a useful and widelyimplemented diagnostic and research tool. Thetechnique involves wedging the flexible broncho-scope into a subsegmental bronchus, instillingsterile saline solution, and retrieving as much fluidas possible back through the channel using suction.Unfortunately, such a simple and widely acceptedtechnique is not fully standardized and certainaspects—such as the amount of fluid injected, thenumber of aliquots used, or the technique forapplying suction—can vary greatly from hospital tohospital. There have been attempts to obtain somestandardization, mainly by way of expert consen-sus.1–3 As far as the authors know, only onerandomized trial, comparing two syringe sizes forBAL, has been published in English.4 Some hospitalsin Catalonia (an area with a population of 6.5million in northeastern Spain) empirically intro-duced a technical modification in the instillationand recovery process consisting in attaching plastictubing between the syringe and the workingchannel of the flexible bronchoscope.

The aim of our study was to test the effect of thistechnical modification in the instillation and recov-ery process. To that end, a multicenter randomizedstudy was designed to be carried out in 11 hospitals.The primary objective was to test whether thisvariation could increase the amount of recoveredfluid, improve the diagnostic yield and decrease thenumber of complications in comparison with theconventional BAL technique without tubing. Asecondary objective was to determine if this methodcould increase the yield specifically for segments notlocated in the anterior part of the lungs.

Patients

BAL was performed in consecutive adult patientsreferred to a bronchoscopy unit to whom a BAL was

indicated. The patients were randomly assigned tothe tubing group or the conventional (no tubing)group following sequences of ten binary randomnumbers. Each center performed both methods.Contraindications for BAL included lack of patientcooperation, hypercapnia and hypoxemia uncor-rected by supplemental oxygen administration,serious cardiac arrhythmias, recent acute myocar-dial infarct or recent asthma attack, uncorrectedbleeding diatheses, and hemodynamic instability.

BAL was performed with tubing in 155 subjectsand without tubing in 140 patients. No differencesbetween groups were found for patient character-istics, indications, lavaged sites, or studies per-formed (Table 1).

Method

All bronchoscopy units used a 40 cm long section ofstandard intravenous plastic tubing with an ex-ternal diameter of 4mm and an internal diameterof 3mm. This section included the rubber portionof the intravenous administration set to which the50mL syringe was inserted and a nude plastic endthat was anchored a few millimeters into theworking channel after its valve was removed(Fig. 1). BAL was performed with 150mL except ifthe total amount recovered after the secondsyringe was less than 10mL. Premedication withintravenous midazolam was optional. Local an-esthesia was provided by application of lidocaine.

Patients were initially placed in a semirecumbentposition (301), supplementary oxygen was adminis-tered through a facial mask or nasal prongs andoxygenation status was monitored during theprocedure with a pulse oximeter (SpO2). Significantoxygen desaturation was considered if SpO2 stayedbelow 90% for more than 30 s. If prolongeddesaturation below 90% occurred, then a higherFiO2 was administered and bronchoscopy wasusually stopped.

ARTICLE IN PRESS

Figure 1

Table 1 Patient characteristics.

Without tubing (n ¼ 155) With tubing (n ¼ 140) P

Gender (male) 101 (65%) 89 (63.6%) 0.435Age (SD) 52.8 (17) 55.2 (16) 0.232Current smoker 57 (40.7%) 57 (36.8%) 0.488Pack-years (SD) 35.5 (20) 37.6 (23) 0.592Intravenous sedation 55 (39.3%) 53 (34.2%) 0.365Initial Sp O2 (SD) 96.07 (2.54) 96.15 (2.61) 0.812RUL/ML/RLL (%) 35/51/14 28.2/53.3/18.2 0.498LUL/Lingula/ LLL (%) 19.6/37.2/43.2 11.5/22.9/65.4 0.262Chest X-ray 0.099Interstitial pattern 37 (23.9%) 54 (38.6%)Node or mass 37 (23.9%) 22 (15.7%)Alveolar opacity 56 (36.1%) 43 (30.7%)Other 22 (14.1%) 19 (13.6%)

Cytology sample 125 (89.3%) 140 (90.3)% 0.769Microbiology sample 114 (81.4%) 129 (83.2%) 0.686Differential cell count 88 (62.9%) 89 (57.4%) 0.341

RUL indicates right upper lobe; ML, middle lobe; RLL, right lower lobe; LUL, left upper lobe (except lingula); LLL, left lowerlobe; SpO2, arterial oxygen saturation by pulse oximetry.

A new BAL fluid instillation and aspiration technique 531

Different models of flexible bronchoscope wereused, with outer diameters ranging from 4.8 to6.2mm and inner diameters ranging from 2.0 to2.8mm. The instrument was inserted transnasallyas the first option, and in case of failure, oralaccess was attempted. The lavage fluid used wassterile pyrogen-free isotonic saline at room tem-perature; buffered solutions were not used. In-stillation and suction was performed manually bydedicated nurses. Dwell time before suction of fluidwas kept to a minimum and patients breathed attidal volume. The decision as to which lobe tolavage, how to handle and process BAL dependedexclusively on the usual practice in each hospital,based on assessment of individual patient status.

Lavage fluid was not filtered before processing. Theliquid recovered after instillation of the firstaliquot was not discarded, so that all the fluidcould be pooled and analyzed as a single BALsample. The total cell count was obtained using aNeubauer’s chamber and cell viability was assessedby Trypan Blue dye exclusion. Alveolitis was definedas any abnormal population cell count (lympho-cytes X15%, neutrophils X5%, eosinophils X2%).Failure was defined as recovery of less than 10mL.5

Only immediate complications were recorded. Ifwedging of the bronchoscope was incomplete,the patient was removed from the study. Eachcenter analyzed the samples in their respectivelaboratories.

Microbiology studies included quantitative bac-terial culture (positive if an organism was isolatedin numbers more than 1000 colony-forming units/mL), fungal smear and culture, viral culture, directfluorescent antibody for Pneumocystis carinii pneu-monia (PCP), and mycobacterial studies. Cytome-galovirus was considered to be a pathogen if grownvia shell culture from the BAL sample or if inclusionbodies were present on histopathological evalua-tion. PCP was diagnosed by a positive directimmunofluorescence antibody or Gomori methyn-damine silver stain. For cytology studies, BAL wascytospun and then stained by the Papanicolaoumethod.

The results from other bronchoscopic proceduresthat might have been used according to thepulmonologist’s criteria in each case were notpooled with the BAL results.

ARTICLE IN PRESS

A. Rosell et al.532

Statistical analysis

SPSS 10.0 (Chicago, IL, USA) was used for statisticalanalysis. Values were expressed as means forquantitative variables and as absolute and relativefrequencies for categorical variables. Total fluidvolume recovered was considered the main outcomevariable; other dependent variables were finaldiagnoses, number of failures, and immediate com-plications. Age, gender, current smoking (X1 cigar-ette per day), smoking index (pack-years), oxygensaturation, chest X-ray findings, use of intravenoussedation, lavaged lung segment and studies per-formed were considered independent variables. First,an analysis was performed to determine whether thetwo groups had similar characteristics, using aStudent t-test for continuous variables and a w2-testfor categorical variables. Then, the outcome vari-ables (total volume, number of diagnoses, failuresand complications) were compared. Results wereconsidered statistically significant with a P value lessthan or equal to 0.05 (two-sided).

Results

The mean total amount of fluid recovered in the groupthat used tubing was 64.8mL (SD 24.4mL, 43.2% of150mL) whereas 52.8mL (SD 27.8mL, 35.2% of 150mL)was obtained in the group without tubing. The 12mLdifference was statistically significant (Po0:005;confidence interval [CI] 95%, 5.9–18.0mL), as wasthe difference in percentage of instilled fluid recov-ered (8%; 95% CI 4.6%–13.5%, Po0:005). In all thethree aliquots there were statistically significantdifferences in the volume recovered (Fig. 2). In 62patients from the only hospital where cell viability wasable to be assessed, non-statistically significantdifferences were found between the two groups(72.7% viable in the group without tubing and 83%viable in the group with tubing, Po0:05).

Figure 2

The tubing technique led to 17.4% more diag-noses, (95% CI 5.9%–28%; P ¼ 0:002) (Table 2). Inboth groups, patients for whom a final diagnosiswas achieved had a greater amount of recoveredfluid. In the group without tubing, the mean volumeof diagnostic BAL recovered was 57.4 cm3 (range,15–115 cm3; SD 23.4 cm3) whereas the mean volumeof non-diagnostic BAL was 49.5 cm3 (range,3–110 cm3; SD 30.1 cm3). In the group with tubing,the mean volume of diagnostic BAL was 68.6 cm3

(range, 20–110 cm3; SD 23.6 cm3) whereas the meanvolume of non-diagnostic BAL was 59.2 cm3 (range,4–105 cm3; SD 24.7 cm3).

More complications occurred in the group with-out tubing than in the group with tubing (8.3% vs.1.4%; P ¼ 0:002) for a difference between groups of6.9% (95% CI 2.0%–12.5%). Oxygen saturation at theend of the procedure tended to be slightly lower inthe group without tubing (94.2%) than in the groupwith tubing (94.7%), but the difference was notstatistically significant. Nevertheless, for thosesubjects in whom desaturation was described as acomplication, mean initial SpO2 was 95.70% andfinal SpO2 was 88.20% (P ¼ 0:004). Recovery fail-ures were also more frequent in the group withouttubing (10.9%) than in the group with tubing (2.1%)being the difference between groups 8.8%, 95% CI3.2%–14.9%; P ¼ 0:003. The most important com-plications declared—respiratory arrest, severebronchospasm and neurological alteration (dizzi-ness)—were three cases of technical failure.

The volume obtained from the anterior areas ofthe lung (anterior segments from right upper lung,left upper lung, lingula, and middle lobe) wasgreater than the volume recovered from othersegments, both with tubing (72.2mL, SD 21.6, vs.52.5mL, SD 24.1; Po0:005) and without tubing(60.8mL, SD 25.3, vs. 41.0mL, SD 27.2; Po0:005).

A questionnaire with closed answers was passedto nurses after the trial was finished. All the nursesreported that they preferred performing BAL withtubing because a more comfortable arm positionwas possible, allowing better control of the syringein case of intense coughing, sneezing, or othermovements of the patient’s head and torso. Otherreasons nurses mentioned for preferring to usetubing were that they felt they had better controlover the negative pressure applied and that it waseasier to change syringes.

Discussion

This randomized controlled clinical trial found thatconnecting plastic tubing between a 50mL syringe

ARTICLE IN PRESS

Table 2 Diagnoses and complications.

Without tubing n ¼ 155 With tubing n ¼ 140 P

DiagnosesTuberculosis 8 (5.2%) 12 (8.5%)Pneumonia (viral, Pseudomonas aeruginosa) 4 (2.6%) 3 (2.1%)Pneumocystis carinii pneumonia 13 (8.3%) 8 (5.7%)Pulmonary hemorrhage 1 (0.6%) 4 (2.8%)Aspergillosis 2 (1.2%) 1 (0.7%)Adenocarcinoma 4 (2.6%) 3 (2.1%)Other 0 2 (1.4%)Alveolitis 33 (21.3%) 50 (35.7%)TOTAL 65 (41.9%) 83 (59.3%) 0.002

ComplicationsAcute respiratory failure 1 (0.6%) 0Bleeding (430mL) 1 (0.6%) 0Bronchospasm 1 (0.6%) 0Dizziness and nausea 1 (0.6%) 0Hypertension and diplopia 0 1 (0.7%)Oxygen desaturation (o90%) 9 (5.8%) 1 (0.7%)TOTAL 13 (8.3%) 2 (1.4%) 0.02

Other included 1 metastasis of breast cancer and 1 eosinophilic granuloma.Alveolitis was defined as any abnormal population cell count (lymphocytesX15%, neutrophilsX5%, eosinophilsX2%) without aspecific diagnoses.

A new BAL fluid instillation and aspiration technique 533

and the working channel of the flexible broncho-scope during instillation and suction facilitatedrecovery of 8% more BAL fluid. The increase wasboth statistically significant and clinically relevantas it was accompanied by more diagnoses and fewerfailures and complications.

A higher percentage of BAL fluid return wasassociated with more final diagnoses. In our study,the lowest volumes that gave a diagnosis were15mL in the group without tubing and 20mL in thestudy group. Because initially recovered BAL fluidcorresponds to a ‘‘bronchial’’ fraction as confirmedby cellular profile6 and radiology,7 a return lessthan 10mL would be unlikely to contain anadequate alveolar sample5 and a procedural failuremight have to be declared. Although the impact ofnegative results in BAL specimens has not beenassessed in our study, it should be remembered thatthe clinical value of BAL lies in the aid they providein ruling out such diseases as tuberculosis, forexample. For a negative result to be useful,however, clinicians must be sure that the sampleis optimal and its volume is adequate.

Different methods for the instillation and recov-ery of saline have been described in the EuropeanRespiratory Society (ERS) and the American Thor-acic Society consensus statements.1,8 The threecommon recovery methods are manual suction,drainage by gravity, and mechanical suction at a

moderate negative pressure (50–100mmHg), but nosingle method is expressly recommended becausethere are no comparative studies. Similarly, there isno recommendation of syringe size (20, 50, or60mL). The only randomized trial focusing on thatissue enrolled 30 patients with chronic obstructivepulmonary disease (COPD) and concluded thatperforming BAL with 50mL syringes obtains morefluid with less oxygen desaturation than can beobtained using 20mL syringes.4 Regarding thevolume to be instilled, the ERS Task Force recom-mends the use of large volumes of at least 100mLto minimize the effect of bronchial contamination,9

but there is still no strong evidence concerning thetotal amount of saline that ought to be instilled.

The percentages of fluid recovered in our study(35.2% without tubing and 43.2% with tubing) weresimilar to those presented in the literature forpatients with lung diseases10,11 but different fromthe recovery of nearly 70% from healthy volun-teers.2 However, once again, the different BALprotocols used and the different populationsstudied make any attempt to establish a strictcomparison between series rather speculative.

The overall complication rate of BAL in theliterature is reported to range between 0% and2.3%, and complications mainly consist of atransient decrease in lung function parameters,alveolar infiltrates, fever, bronchial hyperactivity,

ARTICLE IN PRESS

A. Rosell et al.534

and bronchospasm.11–14 In our patients, there was aclear difference in complication rates betweengroups (8.3% without tubing vs. 1.4% with tubing)that seems to be related to the mean volume ofsaline retained (52.0mL vs. 29.2mL). With respectto complications in the form of lung functionchanges, their extent and duration are stronglycorrelated to the amount of lavage fluid used, tothe amount recovered, and to the underlyingbronchial status of the patients.4 Satisfactory fluidreturn usually consists of removal of more than 50%of the saline solution instilled. Certain conditionssuch as the lavage of lower lobes5 or lower FEV1%

15

will necessarily affect the amount of fluid recov-ered negatively, but any improvement in therecovery rate that can be effected by improvingthe technique would make that technical modifica-tion worthwhile in daily practice.

Because of bronchial lobe anatomy and gravity,returns from lavage in anterior segments aregreater, especially from the middle lobe andlingula.9,12,16 The use of tubing permits a higherreturn of fluid but differences in total fluidrecovery between anterior areas of the lung andthe rest remain similar regardless of technique. Theuse of tubing, therefore, cannot correct the effectof variables related to anatomy and the force ofgravity.

Osteoarticular lesions involving physicians’ ornurses’ necks, shoulders, elbows, and wrists arenot infrequent in bronchoscopy units,16 and smalltechnical modifications in working procedure, suchas the one we tested, can help prevent healthproblems.

It was not the aim of our study to inquire aboutthe biophysical principles that might explain theability of tubing to increase BAL fluid recovery.Nevertheless, it can be hypothesized that thepresence of a semielastic tube between the syringeand the bronchi would attenuate the high negativepeak pressures derived from hand suction; this inturn would influence fluid dynamics by partiallytransforming turbulent flow into laminar flow. Anexperimental study would be required to assess thishypothesis.

In summary, this multicenter randomized studydemonstrates that performing BAL with a tubeplaced between the working channel and a 50mLsyringe increases the volume of fluid recovered bysuction. The increase is statistically significant andclinically relevant as more diagnoses are achieved,fewer complications are reported, and fewerprocedural failures occur. Nevertheless, this mod-ification does not seem to solve the problem of lesssuccessful recovery from the posterior areas of thelung. The system was not tested against other

methods, such as gravity drainage or mechanicalsuction, so it remains unknown how those techni-ques would compare to the tubing one. However,given these results, we recommend this easy-to-apply modification in BAL procedures to thosepulmonologists and nurses already using a handheldsyringe method for recovery. More comparative,randomized, controlled trials are needed to defineoptimal techniques in BAL procedures.

Acknowledgments

The authors would like to thank the SocietatCatalana de Pneumologia (SOCAP) for its constantsupport to the group, Mary Ellen Kerans for hereditorial help in preparing the manuscript, all thenurses from the bronchoscopy units, and SandraAlonso and Carles Farro for their excellent secre-tarial work.

References

1. Klech H, Pohl W. Technical recommendations and guidelinesfor bronchoalveolar lavage (BAL). Report of the SEP TaskGroup on BAL. Eur Respir J 1989;2:561–85.

2. The BAL Cooperative Group Steering Committee. Bronch-oalveolar lavage constituents in healthy individuals, idio-pathic pulmonary fibrosis, and selected comparison groups.Am Rev Respir Dis 1990;141:S169–202.

3. Goldstein RA, Rohatgi PK, Bergofsky EH, Block ER. Clinicalrole of bronchoalveolar lavage in adults with pulmonarydisease: official ATS statement. Am Rev Respir Dis 1990;142:481–6.

4. Blasio F, Rotondetto S, Sarno M, Pezza A. Arterial oxygendesaturation as a consequence of different bronchoalveolartechniques. J Bronchol 1995;2:107–12.

5. Baughman RP, Rennard SI. Bronchoalveolar lavage: generalapproaches to correct for variability of dilution and lungpermeability. Eur Respir J 1999;9:28–31.

6. Rennard SI, Ghafouri M, Thompson AB, et al. Fractionalprocessing of sequential bronchoalveolar lavage to separatebronchial and alveolar samples. Am Rev Respir Dis 1990;141:208–17.

7. Kelly CA, Kotre CJ, Ward C, Hendrick DJ, Walters EH.Anatomical distribution of bronchoalveolar lavage fluid asassessed by digital subtraction radiography. Thorax 1987;42:624–8.

8. Reynolds HY. State of the art: bronchoalveolar lavage. AmRev Respir Dis 1987;135:250–67.

9. Rennard SI, Aalbers R, Blecker E, lech H, Rosenwasser L,Olivieri D, Sibille Y. Bronchoalveolar lavage: performance,sampling procedure, processing and assessment. Eur Respir J1998(Suppl. 26):13s–5s.

10. Finley TN, Swenson EW, Curran WS, Huber GL, Ladman AJ.Bronchopulmonary lavage in normal subjects and patientswith obstructive lung disease. Ann Intern Med 1977;66:651–8.

11. Pingleton AK, Harrison GF, Stechschulte DJ, Wesselius LJ,Kerby GR, Ruth WE. Effect of location, pH and temperature

ARTICLE IN PRESS

A new BAL fluid instillation and aspiration technique 535

of instillate in bronchoalveolar lavage in normal volunteers.Am Rev Respir Dis 1983;1128:1035–7.

12. Ettensohn DB, Jankowski MJ, Duncan PG, Lalor PA. Bronch-oalveolar lavage in the normal volunteer subject. Safety andresults of repeated BAL, an its use In the assessment ofintrasubject variability. Chest 1998;94:281–5.

13. Burns DM, Shure D, Francoz R, Kalafer M, Harrell J, WitztumK, Moser KM. The physiological consequences of saline lobarlavage in healthy human adults. Am Rev Respir Dis 1983;127:695–701.

14. Klech H, Pohl W, Hutter C. Safety and side-effects ofbronchoalveolar lavage. Eur Respir Rev 1992;2(review 8):54–7.

15. Martin TR, Raghu G, Maunder RJ, Springmeyer SC. The effectof chronic bronchitis and chronic air flow obstruction on lungcell population recovered by BAL. Am Rev Respir Dis 1985;132:254–60.

16. Rosell A, De la Rosa D, Badosa R. Lesions in health workers ofbronchoscopy units. [Abstract] Arch Bronchoneumol 2004;40(Suppl. 2):1–151.