General Surgery

THE CONCEPT

A wound is any break in the normal structure and continuity of a tissue. Wound healing is the body's response to that injury — a remarkably organised, overlapping cascade of cellular and biochemical events whose goal is to restore the barrier and, as far as possible, the function of the tissue. It is important to understand that these 'phases' are not separate boxes that finish one after another; they overlap continuously, and a problem in an early phase (say, ongoing infection prolonging inflammation) holds up every later phase. Understanding the sequence explains almost everything about why wounds heal well or badly.

PHASE 1 — HAEMOSTASIS (immediate)

The moment a vessel is cut, it constricts to limit blood loss, and platelets stick to the exposed collagen of the damaged wall, aggregating into a platelet plug. The coagulation cascade then lays down fibrin, converting the plug into a stable clot. This clot does two jobs at once: it stops the bleeding, and it forms a temporary scaffold along which the repair cells will later migrate. Crucially, activated platelets also release growth factors (PDGF, TGF-β) that summon inflammatory cells — so haemostasis is not merely 'stopping bleeding', it is the trigger that starts the whole healing programme.

PHASE 2 — INFLAMMATION (0–3 days)

Neutrophils arrive first, over the first day or two, to kill bacteria and clear debris. From about day 2–3 the macrophage takes over and becomes the key orchestrating cell of healing: it phagocytoses dead tissue and bacteria, and secretes the growth factors (PDGF, TGF-β, VEGF, FGF) that call in fibroblasts and drive new blood-vessel formation. This is why inflammation is a double-edged sword — too little (as in immunosuppression or steroid use) means poor debridement and weak signalling, while too much or too prolonged (as with persistent infection or a foreign body) traps the wound in this phase and prevents it moving on.

PHASE 3 — PROLIFERATION (3–21 days)

Now the wound is rebuilt. Fibroblasts migrate in and synthesise collagen (initially the weaker type III) and ground substance. At the same time, new capillaries sprout (angiogenesis), and together the new vessels and collagen form granulation tissue — the red, granular, slightly bleeding tissue seen in an open healing wound; its redness is simply the many capillary loops, and healthy granulation is the essential bed on which a wound epithelialises or accepts a skin graft. Epithelial cells then migrate across the surface from the wound edges and any surviving skin appendages (epithelialisation), and specialised contractile fibroblasts (myofibroblasts) draw the wound edges together (contraction), shrinking the defect.

PHASE 4 — MATURATION / REMODELLING (3 weeks – up to 1 year)

The wound is now closed but weak. Over months, the disorganised type III collagen is progressively replaced by stronger, cross-linked type I collagen, and the fibres reorganise along the lines of mechanical stress. This is why a wound keeps gaining strength long after it looks healed — yet even a fully mature scar regains only about 70–80% of the original tissue's tensile strength. The scar also becomes paler and flatter as its excess capillaries regress.

TYPES OF WOUND HEALING

How a wound heals depends chiefly on how much tissue is lost and whether the edges can be brought together:

The logic is simple: apposed edges (primary) have almost no gap to bridge, so healing is quick with a neat scar; an open wound (secondary) must laboriously fill a defect with granulation and pull itself closed, which takes longer and scars more.

FACTORS THAT IMPAIR HEALING — AND WHY

Local factors act right at the wound: infection prolongs inflammation and consumes resources; a foreign body is both a mechanical barrier and a nidus for infection; ischaemia/poor blood supply starves the wound of the oxygen collagen synthesis needs; tension or movement repeatedly disrupts the fragile new tissue; a haematoma separates the edges and is an ideal culture medium; and previous irradiation damages the local fibroblasts and blood vessels.

Systemic factors act through the whole patient: malnutrition deprives the wound of the protein needed for collagen and of key cofactors (vitamin C for cross-linking, zinc for enzymes); diabetes mellitus impairs healing on several fronts (microangiopathy, poor perfusion, neutrophil dysfunction, and high glucose feeding bacteria); corticosteroids and immunosuppression blunt the inflammatory phase and directly suppress collagen synthesis (an effect partly reversed by vitamin A); and smoking, anaemia, advanced age, malignancy, jaundice and uraemia all reduce the wound's capacity to repair.

COMPLICATIONS OF HEALING

Healing can go wrong by being excessive — producing a hypertrophic scar or keloid — or deficient — producing a chronic non-healing ulcer or wound dehiscence. Secondary healing across a joint can cause a disabling contracture, and a long-standing scar or chronic ulcer can rarely undergo malignant change into a Marjolin's ulcer (a slow-growing squamous cell carcinoma).

• Four overlapping phases: haemostasis (clot + growth factors) → inflammation (macrophage is key) → proliferation (granulation, collagen III, epithelialisation, contraction) → remodelling (collagen III→I, strength rises to ~80% over months).
• Healing types: primary (apposed, fine scar), secondary (open, granulation + contraction, more scar), tertiary (delayed primary closure).
• Collagen synthesis needs oxygen + vitamin C — explains why ischaemia, smoking, anaemia and scurvy impair healing.
• Local factors: infection, foreign body, ischaemia, tension, haematoma; systemic: malnutrition, diabetes, steroids, smoking.
• Complications: excess (keloid/hypertrophic), deficient (ulcer/dehiscence), contracture, Marjolin's ulcer.

THE CONCEPT

A surgical site infection (SSI) is an infection that develops at or near a surgical incision within 30 days of the operation (or within 90 days if a prosthesis or implant is left in place). It is one of the commonest hospital-acquired infections, and it matters because it causes pain, prolongs hospital stay, can lead to wound breakdown or incisional hernia, and occasionally progresses to life-threatening sepsis. Almost everything about preventing SSI follows from one idea: infection occurs when the load and virulence of contaminating bacteria overwhelm the patient's local defences. Every preventive measure either reduces the bacterial load or strengthens the host — so if you understand that balance, you can reason out the whole management.

HOW IT DEVELOPS

Every surgical wound is contaminated to some degree, but infection only follows when the bacterial inoculum is large enough (classically around 10⁵ organisms per gram of tissue) or when host resistance is lowered. A foreign body or dead tissue dramatically lowers the number of bacteria needed to establish infection, because it shelters organisms from the immune system — which is why implants, sutures and devitalised tissue are such important risk factors. The infection typically declares itself around the 5th–7th postoperative day, and the commonest organism is Staphylococcus aureus.

CDC CLASSIFICATION (by depth)

The CDC classifies SSI by the anatomical depth involved, which guides how it is managed:

SURGICAL WOUND CLASSIFICATION (predicts risk)

Wounds are graded by how contaminated they are at operation — and the expected infection rate rises accordingly, because a dirtier field means a higher bacterial load:

This classification is practically useful: it tells you which operations genuinely benefit from antibiotic prophylaxis and which patients need especially careful technique.

RISK FACTORS

Patient factors that lower resistance include diabetes, obesity (fat is poorly perfused and prone to infection), malnutrition, smoking, immunosuppression, and any distant untreated infection. Operative factors that raise the bacterial burden or reduce local defence include prolonged surgery, poor haemostasis or rough technique, contamination, a foreign body/implant, and intra-operative hypothermia (which impairs the neutrophils that fight infection).

PREVENTION — THE REASONING BEHIND EACH STEP

Because SSI is a battle between bacterial load and host defence, prevention attacks both sides:

• Optimise the patient — control blood glucose, correct malnutrition and anaemia, and stop smoking; each strengthens the host's ability to resist infection.
• Remove hair by clipping, not shaving — shaving with a razor causes microscopic cuts that quickly become colonised with bacteria, so shaved patients have higher infection rates than clipped ones; this is a classic, counter-intuitive exam point.
• Skin antisepsis — a chlorhexidine or povidone-iodine prep reduces the resident skin flora before incision.
• Antibiotic prophylaxis timed within 60 minutes before incision — the aim is a high tissue drug level at the moment of incision, when contamination occurs; given too early or after the incision it is far less effective. It is a single dose (re-dosed only for long operations or major blood loss), not a prolonged course, because the goal is to cover the operative window, not to treat an established infection.
• Maintain normothermia and good oxygenation intra-operatively — neutrophil bacterial killing is an oxygen-dependent process, so a warm, well-oxygenated patient fights contamination better.

MANAGEMENT OF AN ESTABLISHED SSI

The principle mirrors that of any pus collection: open and drain the wound to release the pus and reduce the bacterial load, send a specimen for culture and sensitivity, and debride any dead tissue. The wound is then managed with regular dressings to heal by secondary intention. Antibiotics are given for spreading cellulitis or systemic sepsis, guided by culture — but, as with an abscess, they are an adjunct and cannot replace drainage of a collection.

• Prophylaxis — a single IV dose of an appropriate agent (e.g. a cephalosporin; add metronidazole for colorectal surgery) within 60 minutes before incision.
• Established infection — culture-guided antibiotics; empirical anti-staphylococcal cover pending sensitivities.

• SSI = infection at the surgical site within 30 days (90 days with an implant); commonest organism S. aureus, usually presents day 5–7.
• Infection = bacterial load/virulence overwhelming host defence; a foreign body/dead tissue lowers the inoculum needed.
• CDC depth: superficial incisional / deep incisional / organ-space. Wound classes clean→dirty predict rising risk.
• Prevent by attacking both sides: optimise patient (glucose, nutrition, smoking); clip don't shave; skin antisepsis; prophylactic antibiotic within 60 min; normothermia/oxygenation.
• Treat established SSI by opening/draining + culture-guided antibiotics; antibiotics can't replace drainage.

WHY THESE ARE FEARED

Both conditions are rapidly progressive, life-threatening soft-tissue infections that can kill within hours to days. The single most important lesson is that the visible skin changes badly underestimate the true extent of the disease underneath — so the surgeon who waits for obvious skin necrosis before operating is already too late. The governing principle is 'time is tissue': survival depends on recognising the infection early and taking the patient to theatre for radical surgery, because antibiotics alone cannot reach or remove dead, avascular tissue.

NECROTISING FASCIITIS — the mechanism

Necrotising fasciitis is infection that spreads along the fascial planes — a relatively avascular, low-resistance layer that offers little barrier, so the infection races along it far faster than the overlying skin reveals. As it spreads, it thromboses the small perforating vessels that supply the skin from below; the skin, cut off from its blood supply, becomes dusky, then blisters (haemorrhagic bullae) and finally necroses. This also explains the hallmark early sign — pain out of all proportion to the visible findings — because the disease under the surface is far more extensive than the modest redness on top.

It is usefully divided by microbiology: Type I is polymicrobial (a synergistic mix of aerobes and anaerobes, typically in diabetics and the elderly, where the enzymes of one organism assist another), while Type II is monomicrobial, classically due to Group A Streptococcus (Streptococcus pyogenes). Fournier's gangrene is the name given to necrotising fasciitis of the perineum and scrotum.

CLINICAL FEATURES

• Severe pain out of proportion to the visible signs — the earliest and most valuable clue.
• Oedema spreading beyond the margin of erythema, dusky discoloured skin, haemorrhagic bullae, and sometimes crepitus (gas in the tissues).
• Marked systemic toxicity — fever, tachycardia, hypotension and rapid clinical deterioration, out of keeping with the modest skin appearance.

MANAGEMENT — WHY SURGERY, NOT ANTIBIOTICS, COMES FIRST

The definitive treatment is urgent, radical surgical debridement of all necrotic tissue, repeated as often as needed ('re-look' operations). The reason antibiotics cannot do this job is that the infected tissue is dead and thrombosed — it has no blood supply, so no drug can reach the bacteria within it; only the knife removes the source. Around this, the patient needs broad-spectrum IV antibiotics covering Gram-positives, Gram-negatives and anaerobes, with clindamycin added specifically to switch off bacterial toxin production, plus aggressive resuscitation and organ support in intensive care and correction of the underlying condition (e.g. diabetes).

GAS GANGRENE (Clostridial Myonecrosis)

Gas gangrene is a fulminant infection of muscle caused by Clostridium perfringens, a spore-forming anaerobe that thrives in deep, contaminated, poorly-oxygenated wounds. Its alpha-toxin (a lecithinase) destroys cell membranes, killing muscle and releasing gas — felt as crepitus — while the toxins cause profound, often fatal, toxaemia. The wound is intensely painful, oedematous and produces a thin foul discharge, with a bronze discolouration of the overlying skin.

Treatment again rests on urgent debridement or amputation to remove the dead muscle, combined with high-dose penicillin plus clindamycin (clindamycin, again, suppresses toxin synthesis), full resuscitation, and — where available — hyperbaric oxygen, which works because these organisms are anaerobes for which oxygen is toxic.

INVESTIGATIONS — BUT DON'T DELAY SURGERY

Necrotising fasciitis is fundamentally a clinical diagnosis, and the cardinal error is to wait for tests before operating. Investigations support but never replace judgement: markedly raised inflammatory markers (CRP, white cell count), deranged sodium and renal function, and imaging (X-ray or CT) that may show gas tracking along the fascial planes. The LRINEC score (using CRP, total leucocyte count, haemoglobin, sodium, creatinine and glucose) helps stratify risk, but a low score does not exclude the diagnosis. The definitive diagnosis is often made at operation, where the surgeon finds grey, oedematous, non-bleeding fascia that separates easily on a finger sweep, with thin 'dishwater' pus and a lack of the normal healthy resistance and bleeding of living tissue.

DIFFERENTIAL & PROGNOSIS

The main differentials are severe cellulitis and a simple abscess — but pain out of proportion, rapid spread, systemic toxicity and skin necrosis point firmly to a necrotising infection. Prognosis is directly tied to time: mortality rises sharply with every hour of delay to debridement, and is worse with delayed presentation, extensive disease, older age and comorbidity. After the sepsis is controlled by repeated debridements, the patient needs nutritional support and later reconstruction/skin grafting of the large tissue defects created.

• Gas gangrene — high-dose IV benzylpenicillin + clindamycin (clindamycin for its anti-toxin effect).
• Necrotising fasciitis — broad-spectrum (e.g. piperacillin-tazobactam or a carbapenem) + clindamycin, then de-escalate by culture.

• Both are surgical emergencies — the skin signs underestimate the deep extent; 'time is tissue'.
• Necrotising fasciitis spreads along avascular fascia and thromboses skin vessels → dusky skin, bullae; pain out of proportion is the early clue. Type I polymicrobial, Type II Group A Strep; Fournier's = perineum.
• Treat by urgent radical debridement (antibiotics can't reach dead, thrombosed tissue) + broad-spectrum antibiotics + clindamycin + resuscitation.
• Gas gangrene: Clostridium perfringens alpha-toxin → muscle necrosis + gas (crepitus) + toxaemia.
• Gas gangrene: debridement/amputation + penicillin + clindamycin ± hyperbaric oxygen (O₂ is toxic to the anaerobe).

THE CONCEPT

An ulcer is a break in the continuity of the covering epithelium — of skin or of a mucous membrane. Ulcers are extremely common in surgical practice, and the whole art of managing them lies in reading the ulcer to work out its cause, because the treatment of a venous ulcer, an arterial ulcer, a diabetic ulcer and a malignant ulcer are completely different. The clinical appearance — especially the edge — is not just description for its own sake; each feature is a direct clue to the underlying pathology.

CLASSIFICATION

Ulcers can be described in two complementary ways. Clinically, by their current behaviour — a spreading ulcer (actively enlarging, inflamed edge), a healing ulcer (with a sloping, bluish, epithelialising edge), or a callous/chronic ulcer (long-standing, indurated, making no attempt to heal). Aetiologically, by cause — venous, arterial (ischaemic), neuropathic (trophic/diabetic), pressure (decubitus), traumatic, infective (tuberculous, syphilitic), malignant (Marjolin's, squamous or basal cell carcinoma), and tropical.

READING THE EDGE — THE KEY SKILL

The edge tells you what the ulcer is doing, and therefore what is causing it:

So a punched-out ulcer on the tip of a toe with absent pulses is almost certainly ischaemic; the same punched-out ulcer on the sole of a diabetic foot, but painless and surrounded by callus, is neuropathic. The reasoning is the appeal, not rote memory.

UNDERSTANDING THE COMMON ULCERS

Venous ulcers arise from chronic venous hypertension (failing valves): the sustained high pressure damages the skin capillaries in the 'gaiter' area around the medial malleolus, producing the typical shallow, sloping-edged ulcer with surrounding pigmentation and lipodermatosclerosis. Arterial (ischaemic) ulcers come from an inadequate blood supply, so they occur at the most distal, pressure-exposed points (toes, heel), are painful and punched-out, with absent pulses. Neuropathic (diabetic) ulcers result from loss of protective sensation — the patient does not feel repeated minor trauma, so a painless, punched-out ulcer forms at pressure points on the sole, often with surrounding callus.

MARJOLIN'S ULCER — never forget it

Marjolin's ulcer is a squamous cell carcinoma arising in a chronic wound, scar or sinus — classically in an old burn scar or a long-standing ulcer. Because it grows in scar tissue that has no nerves and few lymphatics, it is characteristically painless and slow to metastasise, and its edge becomes everted. This is the single most important reason why the edge of any chronic, non-healing ulcer must be biopsied — to catch malignant change before it is missed.

INVESTIGATIONS

A wound swab identifies infecting organisms, but the crucial test in a chronic ulcer is biopsy of the edge to exclude malignancy. Vascular ulcers need assessment of the arterial supply with Doppler and the ankle-brachial pressure index (ABI) — this must be done before applying compression, since compressing an ischaemic limb is dangerous. Blood glucose is checked for diabetes, and X-ray looks for underlying osteomyelitis.

MANAGEMENT

The overriding principle is to treat the cause, alongside good local wound care — debridement of slough, appropriate dressings and control of infection. The cause-specific treatments follow directly from the pathology: compression therapy for venous ulcers (to reverse venous hypertension), revascularisation for arterial ulcers (to restore blood supply), and offloading plus tight glucose control and infection management for diabetic foot ulcers, with pressure relief and nursing care for decubitus ulcers. Once a clean, healthy granulating bed is achieved, a skin graft or flap may be used to close a large defect, supported by good nutrition.

HOW TO EXAMINE AN ULCER

A methodical examination is what generates the clues above. On inspection, note the site (e.g. gaiter area suggests venous, toe tip suggests arterial, sole suggests neuropathic), size, shape, number, the all-important edge, the floor (its tissue — healthy pink granulation, or slough, or malignant tissue), and the discharge. On palpation, assess the base for induration (a hard base suggests malignancy or chronic fibrosis), tenderness, and fixity to deeper structures, and — crucially — examine the regional lymph nodes and the vascular and neurological status of the limb, since these determine both cause and management.

PRESSURE (DECUBITUS) ULCER

A pressure ulcer forms over a bony prominence (sacrum, heel, ischial tuberosity, greater trochanter) in an immobile, bedridden or unconscious patient. Sustained pressure exceeds the capillary filling pressure, cutting off the blood supply to the skin and causing ischaemic necrosis. The lesson is that prevention — regular repositioning, pressure-relieving mattresses, good nutrition and meticulous skin care — is far easier and more effective than treating an established ulcer.

PRINCIPLES OF MANAGEMENT (summary)

Identify and treat the underlying cause → Optimise host factors — nutrition, glucose, tissue perfusion → Local wound care — debride slough, dress, control infection → Biopsy the edge of any chronic ulcer to exclude malignancy → Definitive cover (skin graft/flap) once a healthy granulating bed is achieved

• Ulcer = break in covering epithelium; the skill is reading the ulcer to find the cause.
• Classify clinically (spreading/healing/callous) and by cause (venous, arterial, neuropathic, pressure, infective, malignant).
• Edge clues: sloping (healing/venous), punched-out (ischaemic/trophic/syphilitic), undermined (TB), everted (SCC), pearly (BCC).
• Marjolin's = SCC in a chronic scar/ulcer — painless, slow to spread; ALWAYS biopsy the edge of a chronic ulcer.
• Assess vascular status (ABI before compression) and glucose; treat the cause — compression (venous), revascularisation (arterial), offloading (diabetic) — plus wound care.

THE CONCEPT

These three terms describe distinct pathological structures that students often confuse, yet the distinction is simple once you picture them. A cyst is a closed sac; a sinus is a blind-ending tract with one opening; and a fistula is a tract connecting two surfaces with two openings. The key to managing sinuses and fistulae is understanding why they refuse to heal — because a tract will keep discharging as long as something is driving it.

CYST

A cyst is a closed sac with a definite epithelial or endothelial lining containing fluid or semisolid material. The lining is what distinguishes a true cyst from a mere fluid collection (a pseudocyst). Cysts are either congenital — such as a dermoid (from ectoderm buried along embryonic fusion lines), a thyroglossal cyst (a midline neck swelling that classically moves with swallowing and tongue protrusion because of its attachment to the hyoid and tongue), or a branchial cyst — or acquired, such as a retention cyst (e.g. a sebaceous cyst, formed when a gland's outlet is blocked), a distension cyst, a cystic tumour, or a parasitic (hydatid) cyst.

SINUS

A sinus is a blind tract, lined by granulation tissue or epithelium, connecting a deep focus of disease to a surface — it has only one opening. The vital point is that a sinus persists because something at its deep end keeps it discharging — dead tissue, a foreign body, or chronic infection. Common examples are a pilonidal sinus (in the natal cleft, driven by ingrown hair — 'jeep-driver's disease' in hairy young men), a tuberculous sinus, chronic osteomyelitis (with a sequestrum of dead bone at the bottom), and a suture/foreign-body sinus. Treat the discharge by removing the cause at the deep end — otherwise it simply recurs.

FISTULA

A fistula is an abnormal communication between two epithelial (or endothelial) surfaces — either between two hollow organs (e.g. a colovesical fistula) or between an organ and the skin (e.g. an enterocutaneous fistula, or a fistula-in-ano). It has two openings. Fistulae may be congenital (e.g. a branchial fistula) or acquired (following surgery, inflammation such as Crohn's disease, malignancy or trauma).

WHY A SINUS OR FISTULA WON'T HEAL — 'FRIENDS'

A tract heals only when the factor sustaining it is removed. The causes of persistence are captured by the mnemonic FRIENDS, and it is worth understanding each rather than merely listing them:

• F — Foreign body (including sutures, dead bone): a constant nidus for infection and discharge.
• R — Radiation: irradiated tissue has poor blood supply and cannot heal.
• I — Infection / Inflammation (e.g. tuberculosis, Crohn's disease): ongoing inflammation keeps the tract open.
• E — Epithelialisation of the tract: once the tract lines itself with epithelium, it behaves like a permanent channel and will not close spontaneously.
• N — Neoplasm: a tumour in the tract prevents healing.
• D — Distal obstruction: if flow beyond the tract is blocked, pressure forces contents out through the fistula, keeping it patent.
• S — Steroids / immunosuppression (and high output): impair healing generally.

INVESTIGATIONS & MANAGEMENT

A sinogram or fistulogram (contrast injected into the opening), or cross-sectional imaging such as MRI for a fistula-in-ano, defines the tract and its connections before any surgery. Management then follows logically: identify and remove the perpetuating factor (foreign body, infection, distal obstruction, treat Crohn's, etc.), and excise the tract (fistulectomy or sinus excision) once the cause is controlled. A high-output fistula (> 500 mL/day, typically a proximal enterocutaneous fistula) additionally demands careful attention to fluid, electrolyte and nutritional support and protection of the surrounding skin — because losing large volumes of intestinal content is itself dangerous.

SEBACEOUS CYST — A COMMON EXAMPLE

A sebaceous (epidermoid/pilar) cyst is a retention cyst that forms when the outlet of a skin gland is blocked, so keratin and sebum accumulate behind it. It presents as a smooth, spherical swelling that is tethered to the skin at a central punctum (the blocked opening) but freely mobile over the deeper tissues. It may become infected, ulcerate, or (rarely) form a fungating mass called Cock's peculiar tumour. Treatment is complete excision of the cyst wall — leaving any wall behind leads to recurrence, because the lining regenerates the cyst.

PILONIDAL SINUS — A COMMON EXAMPLE

A pilonidal sinus is a sinus in the natal cleft that contains hair, seen typically in hirsute young men ('jeep-driver's disease', because prolonged sitting and friction drive broken hairs into the skin). It presents with a discharging sinus, or with an acute pilonidal abscess. Management is to drain any abscess first, then definitively excise the sinus tract with attention to hygiene and hair removal to prevent recurrence — again illustrating the rule that the perpetuating cause (here, hair) must be dealt with.

• Cyst = closed sac with an epithelial/endothelial lining (congenital: dermoid, thyroglossal, branchial; acquired: sebaceous, hydatid).
• Sinus = blind tract, ONE opening (pilonidal, TB, osteomyelitis, foreign body) — persists because of a deep nidus.
• Fistula = abnormal communication between two surfaces, TWO openings (congenital or acquired).
• Persistence follows FRIENDS: Foreign body, Radiation, Infection/Inflammation (Crohn's/TB), Epithelialisation, Neoplasm, Distal obstruction, Steroids.
• Image the tract (sinogram/fistulogram/MRI); remove the perpetuating cause + excise; support fluids/nutrition in high-output fistulae.