Key takeaways

  • A nipple shield is a clinical tool with specific indications (premature baby, flat or inverted nipples, badly damaged nipples in week one, tongue-tie awaiting release, bottle-preference transition), not a default fix for general feeding difficulty.
  • Size matters: the shield tunnel should be about 2 to 3 mm wider than your nipple. Too small compresses and damages the nipple; too large means the baby cannot draw enough nipple to feed well.
  • Shields do not increase supply and can reduce milk transfer per feed, so weight checks, wet-nappy counts and often pumping after feeds are essential while using one.
  • Plan the weaning from day one. Most clinically indicated shield use is weaned off within 2 to 8 weeks once the underlying problem resolves.
  • An IBCLC (lactation consultant) should ideally guide shield choice, fitting and weaning. A pharmacy handing you a shield with a starter kit is not a clinical indication.
  • Long-term shield use is acceptable when weaning repeatedly fails, weight gain is good and supply is maintained. A shielded feed still delivers your milk.

What a nipple shield is, and what it is not

A modern nipple shield is a thin (0.1 to 0.2 mm), flexible food-grade silicone cover that sits over the nipple and most of the areola during a feed. It has a hollow, nipple-shaped tip with a few small holes (usually four to six) through which milk passes from your breast into the baby's mouth. The flat base rests against the areola.

Mechanically, the shield gives the baby a firm, nipple-like shape to latch onto when your own nipple is flat, inverted or too sore to be drawn in. It can also give a familiar feel to a baby who has been bottle-fed and is struggling to take the bare breast. Crucially, the shield does not make or hold milk. Your baby still has to suck hard enough to draw milk through the holes from the breast underneath.

It helps to be clear about what a shield cannot do. It does not increase your supply. It does not permanently fix a deep latch problem (it works around it). It does not treat thrush, mastitis or a bacterial infection, which need their own treatment. And it does not make breastfeeding easier in general, only in specific situations.

The older rubber and latex shields of the 1970s and 1980s caused real supply problems and are obsolete. Today's thin silicone shields are a major improvement but still carry risks if misused. Marketing oversells them: a shield is a cheap, low-margin item, and many Indian pharmacies and baby stores include one in a starter kit with bottles and a pump, with no guidance. The result is a lot of mothers using a shield for situations where it was never needed, only to notice supply or weight problems weeks later.

The position of lactation bodies (ILCA, the Academy of Breastfeeding Medicine, and ILCA's India chapter) is consistent: a nipple shield is a clinical tool with specific indications and a planned exit. Used appropriately, it can rescue a breastfeeding relationship. Used as a default first response, it can undermine feeding that would otherwise have succeeded. The decision to use one, and the plan for using it, should ideally involve an IBCLC.

On materials and care: modern shields are BPA-free silicone, washable with mild soap and warm water after every use, and should be sterilised daily through the first six months (boil for five minutes, or use a microwave or electric steam steriliser). Replace every two to three months, or sooner if you see any tears, holes, discolouration or sticky residue. Never use a latex shield.

When a shield actually helps: the real indications

The Academy of Breastfeeding Medicine and IBCLC guidance describe specific situations where a shield is a sound choice. What they share is a genuine barrier to feeding that positioning, latch correction or simpler measures cannot resolve on their own.

Premature babies (roughly 32 to 36 weeks). A premature baby often cannot generate enough suction to hold a latch on a flat or normal nipple while still learning to feed. A shield gives a firmer structure to latch onto. Research in preterm infants (Meier and colleagues) found that babies actually transferred more milk with a shield than without, because they could maintain the latch. The shield is weaned off as the baby matures, usually by 36 to 38 weeks corrected age. For preterm feeding more broadly, skin-to-skin and kangaroo mother care and getting that first colostrum into the baby matter just as much.

Flat or inverted nipples. Some nipples sit flat against the areola or retract inward when stimulated, and a baby may not be able to draw them in for a deep latch even with good positioning. A shield offers a firm shape to latch onto, and over weeks the baby's suction can gradually draw out the underlying nipple so that direct feeding becomes possible. Many flat nipples respond to positioning and breastfeeding position changes alone, so an assessment first is worthwhile.

Severely damaged nipples in the first week. If cracked, bleeding nipples make direct feeding intolerable in week one and you are close to stopping, a brief spell of shield use (a few days to two weeks) can protect the nipple while it heals and keep feeding going. The latch problem that caused the damage still has to be fixed; the shield only buys time. Healing the skin matters too, which we cover in how to treat cracked nipples. This indication is debated, and some consultants prefer to correct the latch directly rather than add a shield.

Bottle-preference transition. A baby who has had a lot of bottles early (a NICU baby, a baby separated from the mother, or one supplemented by bottle) may refuse the bare breast because the bottle was easier. The firmer, more bottle-like shield can bridge the baby back to the breast, then be weaned off.

Tongue-tie awaiting treatment. A baby with Tongue-Tie & Lip-Tie in Indian Babies: Signs, Frenotomy, Care who cannot make a deep latch may be helped temporarily by a shield while the frenotomy (a quick release of the tight band under the tongue) is arranged. After the release, the shield is weaned off as the deeper latch develops.

Breast surgery or scarring affecting the nipple, and some oral-motor or sensory differences in the baby (for example Down syndrome, or cleft lip without cleft palate) can also be genuine indications, but these need specialist IBCLC assessment.

What is not an indication: general tiredness with breastfeeding, mild soreness, normal newborn fussiness, perceived low supply without assessment, a family member's suggestion, or simply that feeding feels harder than expected. These need lactation support, not a shield as a first response.

The simple test an IBCLC applies: is there a specific problem the shield will address; have simpler steps been tried; is there a weaning plan; will weight and milk transfer be monitored; is the mother able and willing to follow the plan? If yes to all, a shield trial is reasonable.

Sizing and fitting: the detail that decides success

A shield only works if it fits your nipple. A mis-sized shield damages the nipple, cuts milk transfer and frustrates both of you, which is why fitting is the most important part of a shield consultation.

Measure the nipple, not the length. The nipple is measured at its base (where it meets the areola), in its erect state, just before a feed or after brief stimulation. Most mothers measure roughly 14 to 22 mm across, and the two breasts can differ. A lactation consultant can use a simple nipple ruler.

Match the size with breathing room. The shield's internal tunnel should be about 2 to 3 mm wider than your nipple. So roughly: a 15 mm nipple suits a 16 mm shield, an 18 mm nipple a 20 mm shield, a 22 mm nipple a 24 mm shield. Too small (a 16 mm shield on an 18 mm nipple) compresses and damages the nipple. Too large leaves empty space and the baby cannot latch effectively. Sizes are not perfectly standardised across brands, so a Medela 20 mm may feel different from a Pigeon or Chicco 20 mm; trying another brand sometimes solves a poor fit.

Apply it well. Lightly moisten the inner surface with water or breast milk for grip, centre it on the nipple with the base flat against the areola, and turn any cutout (some brands leave a gap so the baby can smell the breast) toward the baby's nose. Then latch the baby as you would on the breast: wide-open mouth, lips flanged out, and the base of the shield (over the areola) drawn well into the mouth, not just the silicone tip. The deep, wide latch still applies. If latching is the core struggle, a structured latch-troubleshooting approach is worth working through.

Read the signs of a good feed versus a poor one. With a good fit, the baby latches deeply, you hear regular swallowing, you may see milk pooling at the shield tip, the feed lasts a normal 15 to 30 minutes, the nipple is not damaged, the breast feels softer afterwards, and nappies and weight gain are on track. Warning signs: the shield slips, lips are tucked in rather than flanged, no audible swallow, no visible milk, feeds dragging out to 45 to 60 minutes, a fussy unsatisfied baby, poor weight gain, breasts as full after as before, or pain during feeding (a sign of compression).

Protecting milk transfer and supply

The biggest concern with a shield is whether the baby is actually getting enough milk. Direct feeding without a shield transfers nearly all the available milk; feeding through a shield can transfer less in some babies. So monitoring is not optional.

What research shows. Results vary. In premature babies, transfer can be higher with a shield because the baby can hold the latch. In some full-term babies, studies (such as Chertok's work) have found meaningfully reduced transfer per feed. How much depends on the baby's strength, the fit, your let-down and the situation. The extra silicone barrier can reduce suction efficiency, the small tip holes can limit flow, and a shallower draw can mean less stimulation to the breast, which over weeks can nudge supply down.

Weight gain is the gold standard. In the first three months, average gain is roughly 20 to 30 g per day (about 140 to 210 g per week). Babies on a shield should have weekly weight checks (with the paediatrician or IBCLC) for the first four to six weeks of use. Gain below the lower end is a reason for urgent reassessment.

Count nappies. From around day six, expect at least six wet nappies in 24 hours (pale, clear urine) and several yellow, seedy, loose stools a day. Fewer wet nappies can signal low intake.

Watch satisfaction. A well-fed baby settles for one and a half to two hours between feeds, wakes alert and feeds eagerly. A baby not getting enough feeds constantly, stays fussy, or is hard to wake and sleepy.

Pump to protect supply. Many consultants advise pumping for 5 to 10 minutes after feeds (or after a few feeds a day), especially in the first four to six weeks, to fully drain the breast and keep the supply signal strong. Practical pumping and milk storage makes this easier to sustain. Any expressed milk can be given by spoon or cup if extra intake is needed.

Maintain supply over weeks. Alongside pumping, eat well (breastfeeding needs roughly 500 extra calories a day) and stay hydrated. Shatavari is widely used in Indian practice and is considered safe, though the evidence for galactagogues is mixed; use them on a consultant's advice rather than as a substitute for drainage. If supply is genuinely dropping, an assessment and a structured plan help, and our guide on how long it takes to rebuild milk supply sets realistic expectations.

If shielded feeds still are not enough despite all this, short-term supplementation with expressed milk (by cup or syringe, not by bottle, to avoid worsening any preference) protects the baby's nutrition. Where formula is medically needed, safe formula preparation matters. The aim is always to return to fully nourishing feeds, shielded or direct, as soon as you can.

Weaning off the shield: the plan and the patience

Decide on weaning at the start, not weeks later when the baby will only feed with the shield. The guidance is simple: attempt weaning at every follow-up once the underlying problem has eased.

When to start. Begin once the original reason improves. For damaged nipples, usually one to two weeks after they heal. For tongue-tie, a few days after the frenotomy. For a premature baby, around 36 to 38 weeks corrected age. For flat or inverted nipples, once the nipple has been drawn out enough to protrude.

Five methods that work, roughly in order of how often they succeed:

Mid-feed removal (highest success). Start with the shield, let the baby latch and feed. After two to three minutes, once let-down is flowing and the baby is calm and rhythmic, slip the shield off and re-latch directly. The flowing milk makes the bare breast more rewarding. If the baby fusses, replace the shield and try again next feed.

Skin-to-skin and laid-back nursing. Undress to skin contact, recline at about 45 degrees, place the baby on your chest and let them self-attach. Instinctive feeding reflexes are strongest here, and many babies latch directly without the shield, especially at a rested morning feed. The wider benefits of skin-to-skin care make this worth doing regardless.

The hungry first feed of the day. The baby is most motivated and coordinated at the first morning feed. If they latch directly then, try the rest of the day without the shield.

Gradual replacement. Swap one shielded feed a day for a direct attempt at a calm, alert feed, then add a second and third as the baby succeeds.

Bait and switch. Start shielded, switch to direct after a minute, switch back if the baby objects, then try direct again. Rapid alternation can win the baby over within one session.

Keep the shield handy while weaning. Do not throw it away the day you start. A hard feed, teething or an illness may mean falling back on it for a few feeds, and having it on hand avoids panic. Once a couple of weeks pass without it, put it away.

Expect it to take time. Weaning often takes two to eight weeks of consistent effort. Some babies wean after one or two good direct feeds; others, especially after prolonged use, take longer. Follow-up every couple of weeks helps with troubleshooting.

If weaning keeps failing, the underlying issue may not be fully resolved (the nipple is still flat, the latch problem persists, the oral-motor difficulty remains). A reassessment can find what is holding the dependency in place. Sometimes continued shield use is genuinely the right call.

When long-term shield use becomes the plan

Most clinically appropriate shield use is short-term, two to eight weeks, ending in direct feeding. A minority of families use a shield long-term, by choice or because weaning has not worked. That is acceptable when feeding is going well.

Why it happens. Very flat or persistently inverted nipples that never fully draw out; persistent oral-motor differences in the baby; a well-established breastfeeding routine the family is comfortable with and chooses not to change; or nipple pain that returns whenever direct feeding is attempted.

Is it safe? There is no specific evidence that long-term shield use harms baby or mother, provided milk transfer and supply stay adequate. The real concerns are keeping supply up over months and the practical hassle of an extra step at every feed (a forgotten or dropped shield, hygiene).

Monitor over the long term. Monthly weight checks after the first six weeks, ongoing attention to nappies and your sense of fullness and let-down, a lactation review every two to three months even when all seems fine, and a fresh shield every two to three months. Wash after every use and sterilise daily for the first six months, then at least weekly.

Support supply over months. Feed frequently (every two to three hours through the first six months, every three to four hours after), keep some night feeds (they support prolactin), pump occasionally after feeds to ensure full drainage, and eat and drink well.

Keep trying gently. Even with long-term use, a soft weaning attempt every two to three months is worth it as the baby's oral skills mature. If you also bottle-feed during work hours, paced bottle feeding (bottle held level, baby controlling the flow, feed taking 15 to 20 minutes) preserves direct-feeding skills.

On the emotions. Long-term shield use is not a failure. The goal is for your baby to receive your milk and for the feeding relationship to thrive, and a shielded feed delivers both. Some mothers feel completely at ease with it; some prefer it. Both are valid. When you do decide to stop breastfeeding entirely, the same principles apply whether feeds were shielded or direct, covered in our guide on stopping breastfeeding without pain.

Try these first: alternatives to a shield

Many situations where a shield is considered would actually resolve with a simpler step. A good consultant works through these first.

Latch correction. Most early difficulty is a latch issue fixable with hands-on coaching: wide-open mouth, lips flanged out, plenty of areola in the mouth, chin pressed into the breast, an asymmetric latch with more areola showing above the top lip than below the chin. Watching and correcting a feed often solves what would otherwise have led to a shield.

Position changes. The cradle hold is the default but not always best. The cross-cradle gives more head control for newborns, the football hold suits caesarean recovery, large breasts and twins, side-lying helps at night, and laid-back with skin contact is excellent for early days. Working through different breastfeeding positions can resolve issues a shield would only mask. For twins specifically, tandem feeding techniques make a real difference.

Reverse pressure softening. If engorgement makes the areola too firm to latch, gentle inward pressure around the nipple base for a minute or two softens it so the baby can latch. This often beats a shield in the early engorged days, and our guide to relieving breast engorgement covers it in detail.

A few drops of hand-expressed milk on the nipple before latching can tempt a reluctant baby with the smell and taste.

Treat the real cause of pain. Thrush, mastitis, a bacterial infection or vasospasm all cause nipple pain that gets blamed on the latch and leads to a needless shield. Treating the actual cause removes the pain. If pain is sharp and deep, read about the causes of shooting breastfeeding pain.

Frenotomy for tongue-tie. If tongue-tie is the problem, the definitive answer is releasing it, after which many babies latch deeply without a shield. Using a shield to work around an untreated tongue-tie only delays the real fix.

Address supply at its source. If low supply is the worry, frequent feeding, pumping after feeds, good nutrition and hydration build supply. A shield does not; it may reduce it. Using a shield as a response to perceived low supply is counterproductive.

Once these have been tried and have not worked, a shield becomes a reasonable, well-guided choice with a weaning plan, neither a first reflex nor a last resort to avoid.

India context: cost, brands and family attitudes

Nipple shields are easy to find across Indian pharmacies, baby stores and online. That access is a benefit (you can get one quickly when you need it) and a risk (you can buy and misuse one without guidance).

Where to buy. Pharmacy chains (Apollo, MedPlus, Wellness Forever), baby stores (FirstCry, Hopscotch), online (Amazon, Flipkart, FirstCry) and neighbourhood chemists. Some hospitals provide a shield at discharge.

Brands and prices. Medela Contact (the international benchmark, roughly Rs 800 to 1,500, in 16, 20 and 24 mm); Pigeon (about Rs 500 to 1,200, sizes S/M/L corresponding to roughly 16, 20 and 24 mm); Philips Avent (about Rs 700 to 1,400, sizes around 15 and 21 mm); Chicco Soft (about Rs 400 to 900, M and L); Mee Mee (about Rs 200 to 500, generic sizing). Cheaper generic shields may have less precise sizing and thicker silicone, so Medela or Pigeon are generally preferred for clinical use. Most Indian health insurance treats shields as consumables and does not cover them, so cost is out of pocket.

Family attitudes. Reactions vary. Some grandmothers recall the old rubber shields, some have never seen the silicone version, and some read it as a sign the mother is not feeding properly. Explaining the clinical reason (a consultant or paediatrician recommended it for a specific issue) and framing it as a temporary tool usually settles the social pressure.

Hospital practice varies widely. Some hospitals have IBCLC-trained nurses who fit shields properly; others hand them out at discharge without instruction; some discourage them. If you were given a shield without explanation, ask for a lactation or paediatric follow-up to check whether it is right for you and to set a plan.

Government and community support. ASHA workers visit postpartum mothers in many states and give basic feeding support, though training on shields specifically varies; for shield questions, an IBCLC or paediatrician is more reliable. The Breastfeeding Promotion Network of India (BPNI) and ILCA's India chapter both treat the shield as a temporary clinical intervention with a clear emphasis: use it only for a real indication, monitor with weekly weight checks for the first four to six weeks, and attempt weaning at every follow-up once the problem resolves.

Tele-lactation. Where in-person consultants are hard to reach, video consultations (typically Rs 1,000 to 2,500) can guide use, fitting and weaning. In-person is best for the first fitting and major troubleshooting; tele works well for follow-up.

Common misuse in India includes using a shield for normal newborn fussiness, months of use without any weaning attempt, sharing a shield between mothers (a hygiene risk), wrong sizing that damages the nipple, and treating the shield as a replacement for lactation support rather than a tool used alongside it.

Working with a lactation consultant (IBCLC)

An IBCLC (International Board Certified Lactation Consultant) is the specialist best placed to guide shield use, from the first fitting through weaning, and their involvement clearly improves outcomes.

Finding one in India. The IBLCE directory (iblce.org) lists verified IBCLCs by country and city, and BPNI maintains an Indian directory. Major cities (Mumbai, Delhi, Bengaluru, Chennai, Hyderabad, Pune, Kolkata) have several in private practice; smaller cities may have only one or two, where tele-consultation is the practical option.

Cost. In-person sessions usually run Rs 1,500 to 3,500, more in metros; tele-consultations Rs 1,000 to 2,500. Some hospital packages include lactation support. It is generally money well spent and can save the feeding relationship.

What to bring to the first visit. A short feeding history (when the baby was born, when feeding started, what has happened since), the baby's weight history from paediatric records, photos of the nipple shape if relevant, any shields you have already tried (with size labels), and your current concerns.

What the consultant assesses. Nipple shape and size, a breast exam, a full feed observed with and without the shield where possible, a weight check, the baby's oral-motor pattern (tongue mobility, palate, suck), and your feeding goals and family situation. From this they set a plan: brand and size, application and latch technique, an expected timeline, a weaning approach for your baby, a monitoring plan, and a follow-up schedule.

Follow-up. Typically one to two weeks for the first month, then every two to four weeks until weaning is complete, reassessing and attempting a weaning trial each time. The IBCLC handles the feeding mechanics; the paediatrician watches the baby's overall health and weight; both are needed, and they work best when information flows between them.

Seek an urgent slot if the baby is not gaining or is losing weight, shows signs of dehydration, feeding is failing, or severe nipple pain is preventing feeds. Many consultants keep emergency slots for exactly these situations.

If no IBCLC is reachable, BPNI and La Leche League India helplines and tele-consultations can guide you, with the paediatrician as the baby's primary medical contact. After successful weaning, a final review confirms the transition held, supply is good and the baby feeds well directly, and the shield can then be cleaned, stored and kept only in case a problem recurs.

When to see a doctor or lactation consultant

A shield is meant to support feeding, not to mask a baby who is not thriving. Seek help promptly if you notice the warning signs below. Weight loss, dehydration and persistent feeding failure are not things to wait out.

Nipple shield myths in Indian families, corrected

Myth: a nipple shield is a simple fix for any breastfeeding problem

  • A shield has specific indications (premature baby, flat or inverted nipples, badly damaged nipples in week one, tongue-tie awaiting release, bottle-preference transition). It is not a default for any difficulty.
  • Many problems resolve with latch correction, position changes or lactation support, with no shield at all.
  • Using a shield without a real indication can create dependency, reduce milk transfer and undermine feeding that would have worked.
  • ILCA, the Academy of Breastfeeding Medicine and ILCA-India all describe the shield as a temporary clinical tool with guidance and a weaning plan, not a first response.
  • A pharmacy or baby store including a shield in a starter kit is not a clinical indication; it may or may not suit your situation.

Myth: once you start a shield, you are stuck with it forever

  • Most clinically indicated shield use is weaned off within two to eight weeks once the underlying issue resolves.
  • Attempting weaning at every follow-up clearly improves success.
  • Methods that work include mid-feed removal, skin-to-skin laid-back nursing, the first morning feed without the shield, and gradually replacing shielded feeds with direct ones.
  • Some babies wean after one or two good direct feeds; others take longer.
  • A small group of families use shields long-term, which is acceptable when monitored, but it is not inevitable.
  • One failed attempt does not mean weaning will never work; patience over weeks is what succeeds.

Myth: a nipple shield increases milk supply

  • Shields do not increase supply and may reduce it over time, because the breast is stimulated less efficiently than with direct feeding.
  • In some full-term babies, feeding through a shield transfers meaningfully less milk per feed than direct feeding.
  • To protect supply during shield use, pump for 5 to 10 minutes after feeds to fully drain the breast.
  • Real supply problems need supply-building steps (frequent feeds, pumping, good nutrition and hydration, galactagogues on advice), not a shield.
  • Weekly weight checks during shield use are the most reliable way to confirm enough intake; gain below about 140 g per week in the first three months warrants reassessment.

Myth: a bigger shield is always better than a smaller one

  • Shield size should match your nipple diameter (measured at the base, erect); the tunnel should be about 2 to 3 mm wider than the nipple.
  • Too small compresses and damages the nipple and causes pain.
  • Too large leaves the baby too little nipple to draw out, reducing milk transfer.
  • Common sizes are 16, 20 and 24 mm; the right one depends on your measurement, ideally taken by an IBCLC.
  • Brands differ, so a Medela 20 mm may feel different from a Pigeon 20 mm; the right brand-and-size combination matters.
  • Re-check sizing over time, as some flat nipples are drawn out by weeks of nursing and may need a larger shield.

Frequently asked questions

Does using a nipple shield reduce my milk supply?

It can, over time, because the breast is stimulated a little less efficiently than with direct feeding, and milk transfer per feed can be lower in some full-term babies. It does not have to reduce supply if you manage it: pump for 5 to 10 minutes after feeds to fully drain the breast, feed frequently, and have weekly weight checks for the first four to six weeks. In premature babies a shield can actually improve transfer by helping them hold the latch.

How do I know which size nipple shield to buy?

Match the shield to your nipple, not the other way around. Measure your nipple at its base in its erect state; the shield's internal tunnel should be about 2 to 3 mm wider. As a rough guide, a 15 mm nipple suits a 16 mm shield, 18 mm suits 20 mm, and 22 mm suits 24 mm. Too small compresses and damages the nipple; too large means the baby cannot latch well. An IBCLC can measure you precisely, and brands vary, so the same number can fit differently.

How long does it take to wean a baby off a nipple shield?

Usually two to eight weeks of consistent effort once the underlying reason for using it has resolved. Some babies wean after one or two successful direct feeds; others, especially after prolonged use, take longer. The most successful method is mid-feed removal: start with the shield, and once let-down is flowing and the baby is feeding calmly, slip the shield off and re-latch directly. Keep the shield handy during weaning in case you need to fall back on it.

Is it safe to use a nipple shield long-term?

Yes, when feeding is going well, weight gain is good and supply is maintained. There is no specific evidence that long-term silicone-shield use harms mother or baby provided milk transfer stays adequate. The main concerns are protecting supply over months and the practical hassle of an extra step. Keep up monthly weight checks, replace the shield every two to three months, and try a gentle weaning attempt every couple of months as the baby's feeding skills mature.

My pharmacy gave me a shield with the starter kit. Should I use it?

Not automatically. A shield included in a starter kit is a marketing choice, not a clinical recommendation for your situation. Using one without a real indication can reduce milk transfer and create dependency. If you are struggling with feeding, get a latch and feeding assessment first, in person or by tele-consultation. A shield is the right answer only for specific problems, and then with proper fitting and a weaning plan.

Can I sterilise and reuse a nipple shield, and how often should I replace it?

Yes. Wash it with mild soap and warm water after every feed, rinse well and air-dry on a clean surface. Sterilise it daily for the first six months by boiling for five minutes or using a microwave or electric steam steriliser, then at least weekly afterwards. Replace it every two to three months, or sooner if you see any tears, holes, discolouration or sticky residue. Never share a shield between mothers, and never use an old latex shield.

Sources